notebooks/pipeline_titanic.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from numpy import interp\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "pd.set_option(\"display.max_columns\", None)\n",
    "\n",
    "from sklearn.compose import ColumnTransformer, make_column_selector\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler, FunctionTransformer, PolynomialFeatures\n",
    "\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.pipeline import FeatureUnion\n",
    "\n",
    "from sklearn.metrics import plot_confusion_matrix, roc_auc_score,  auc, \\\n",
    "    precision_recall_fscore_support, classification_report, roc_curve, plot_roc_curve\n",
    "\n",
    "from sklearn import set_config\n",
    "set_config(display='diagram')   \n",
    "\n",
    "from catboost import CatBoostClassifier\n",
    "\n",
    "from itertools import cycle\n",
    "from time import time\n",
    "\n",
    "from lightgbm import LGBMClassifier"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Import data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train = pd.read_csv('./data/train.csv')\n",
    "df_test = pd.read_csv('./data/test.csv')\n",
    "sample_sub = pd.read_csv('./data/sample_submission.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Oconnor, Frankie</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>209245</td>\n",
       "      <td>27.14</td>\n",
       "      <td>C12239</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Bryan, Drew</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>27323</td>\n",
       "      <td>13.35</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Owens, Kenneth</td>\n",
       "      <td>male</td>\n",
       "      <td>0.33</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>CA 457703</td>\n",
       "      <td>71.29</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Kramer, James</td>\n",
       "      <td>male</td>\n",
       "      <td>19.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>A. 10866</td>\n",
       "      <td>13.04</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Bond, Michael</td>\n",
       "      <td>male</td>\n",
       "      <td>25.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>427635</td>\n",
       "      <td>7.76</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99995</th>\n",
       "      <td>99995</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>Bell, Adele</td>\n",
       "      <td>female</td>\n",
       "      <td>62.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 15008</td>\n",
       "      <td>14.86</td>\n",
       "      <td>D17243</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99996</th>\n",
       "      <td>99996</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>Brown, Herman</td>\n",
       "      <td>male</td>\n",
       "      <td>66.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13273</td>\n",
       "      <td>11.15</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99997</th>\n",
       "      <td>99997</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Childress, Charles</td>\n",
       "      <td>male</td>\n",
       "      <td>37.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9.95</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99998</th>\n",
       "      <td>99998</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Caughlin, Thomas</td>\n",
       "      <td>male</td>\n",
       "      <td>51.00</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>458654</td>\n",
       "      <td>30.92</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99999</th>\n",
       "      <td>99999</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Enciso, Tyler</td>\n",
       "      <td>male</td>\n",
       "      <td>55.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>458074</td>\n",
       "      <td>13.96</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100000 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       PassengerId  Survived  Pclass                Name     Sex    Age  \\\n",
       "0                0         1       1    Oconnor, Frankie    male    NaN   \n",
       "1                1         0       3         Bryan, Drew    male    NaN   \n",
       "2                2         0       3      Owens, Kenneth    male   0.33   \n",
       "3                3         0       3       Kramer, James    male  19.00   \n",
       "4                4         1       3       Bond, Michael    male  25.00   \n",
       "...            ...       ...     ...                 ...     ...    ...   \n",
       "99995        99995         1       2         Bell, Adele  female  62.00   \n",
       "99996        99996         0       2       Brown, Herman    male  66.00   \n",
       "99997        99997         0       3  Childress, Charles    male  37.00   \n",
       "99998        99998         0       3    Caughlin, Thomas    male  51.00   \n",
       "99999        99999         0       3       Enciso, Tyler    male  55.00   \n",
       "\n",
       "       SibSp  Parch     Ticket   Fare   Cabin Embarked  \n",
       "0          2      0     209245  27.14  C12239        S  \n",
       "1          0      0      27323  13.35     NaN        S  \n",
       "2          1      2  CA 457703  71.29     NaN        S  \n",
       "3          0      0   A. 10866  13.04     NaN        S  \n",
       "4          0      0     427635   7.76     NaN        S  \n",
       "...      ...    ...        ...    ...     ...      ...  \n",
       "99995      0      0   PC 15008  14.86  D17243        C  \n",
       "99996      0      0      13273  11.15     NaN        S  \n",
       "99997      0      0        NaN   9.95     NaN        S  \n",
       "99998      0      1     458654  30.92     NaN        S  \n",
       "99999      0      0     458074  13.96     NaN        S  \n",
       "\n",
       "[100000 rows x 12 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100000</td>\n",
       "      <td>3</td>\n",
       "      <td>Holliday, Daniel</td>\n",
       "      <td>male</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>24745</td>\n",
       "      <td>63.01</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100001</td>\n",
       "      <td>3</td>\n",
       "      <td>Nguyen, Lorraine</td>\n",
       "      <td>female</td>\n",
       "      <td>53.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13264</td>\n",
       "      <td>5.81</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100002</td>\n",
       "      <td>1</td>\n",
       "      <td>Harris, Heather</td>\n",
       "      <td>female</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>25990</td>\n",
       "      <td>38.91</td>\n",
       "      <td>B15315</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100003</td>\n",
       "      <td>2</td>\n",
       "      <td>Larsen, Eric</td>\n",
       "      <td>male</td>\n",
       "      <td>25.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>314011</td>\n",
       "      <td>12.93</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100004</td>\n",
       "      <td>1</td>\n",
       "      <td>Cleary, Sarah</td>\n",
       "      <td>female</td>\n",
       "      <td>17.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>26203</td>\n",
       "      <td>26.89</td>\n",
       "      <td>B22515</td>\n",
       "      <td>C</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99995</th>\n",
       "      <td>199995</td>\n",
       "      <td>3</td>\n",
       "      <td>Cash, Cheryle</td>\n",
       "      <td>female</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7686</td>\n",
       "      <td>10.12</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99996</th>\n",
       "      <td>199996</td>\n",
       "      <td>1</td>\n",
       "      <td>Brown, Howard</td>\n",
       "      <td>male</td>\n",
       "      <td>59.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>13004</td>\n",
       "      <td>68.31</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99997</th>\n",
       "      <td>199997</td>\n",
       "      <td>3</td>\n",
       "      <td>Lightfoot, Cameron</td>\n",
       "      <td>male</td>\n",
       "      <td>47.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4383317</td>\n",
       "      <td>10.87</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99998</th>\n",
       "      <td>199998</td>\n",
       "      <td>1</td>\n",
       "      <td>Jacobsen, Margaret</td>\n",
       "      <td>female</td>\n",
       "      <td>49.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>PC 26988</td>\n",
       "      <td>29.68</td>\n",
       "      <td>B20828</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99999</th>\n",
       "      <td>199999</td>\n",
       "      <td>1</td>\n",
       "      <td>Fishback, Joanna</td>\n",
       "      <td>female</td>\n",
       "      <td>41.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>PC 41824</td>\n",
       "      <td>195.41</td>\n",
       "      <td>E13345</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100000 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       PassengerId  Pclass                Name     Sex   Age  SibSp  Parch  \\\n",
       "0           100000       3    Holliday, Daniel    male  19.0      0      0   \n",
       "1           100001       3    Nguyen, Lorraine  female  53.0      0      0   \n",
       "2           100002       1     Harris, Heather  female  19.0      0      0   \n",
       "3           100003       2        Larsen, Eric    male  25.0      0      0   \n",
       "4           100004       1       Cleary, Sarah  female  17.0      0      2   \n",
       "...            ...     ...                 ...     ...   ...    ...    ...   \n",
       "99995       199995       3       Cash, Cheryle  female  27.0      0      0   \n",
       "99996       199996       1       Brown, Howard    male  59.0      1      0   \n",
       "99997       199997       3  Lightfoot, Cameron    male  47.0      0      0   \n",
       "99998       199998       1  Jacobsen, Margaret  female  49.0      1      2   \n",
       "99999       199999       1    Fishback, Joanna  female  41.0      0      2   \n",
       "\n",
       "         Ticket    Fare   Cabin Embarked  \n",
       "0         24745   63.01     NaN        S  \n",
       "1         13264    5.81     NaN        S  \n",
       "2         25990   38.91  B15315        C  \n",
       "3        314011   12.93     NaN        S  \n",
       "4         26203   26.89  B22515        C  \n",
       "...         ...     ...     ...      ...  \n",
       "99995      7686   10.12     NaN        Q  \n",
       "99996     13004   68.31     NaN        S  \n",
       "99997   4383317   10.87     NaN        S  \n",
       "99998  PC 26988   29.68  B20828        C  \n",
       "99999  PC 41824  195.41  E13345        C  \n",
       "\n",
       "[100000 rows x 11 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100001</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100002</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100003</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100004</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99995</th>\n",
       "      <td>199995</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99996</th>\n",
       "      <td>199996</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99997</th>\n",
       "      <td>199997</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99998</th>\n",
       "      <td>199998</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99999</th>\n",
       "      <td>199999</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100000 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       PassengerId  Survived\n",
       "0           100000         1\n",
       "1           100001         1\n",
       "2           100002         1\n",
       "3           100003         1\n",
       "4           100004         1\n",
       "...            ...       ...\n",
       "99995       199995         1\n",
       "99996       199996         1\n",
       "99997       199997         1\n",
       "99998       199998         1\n",
       "99999       199999         1\n",
       "\n",
       "[100000 rows x 2 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_sub"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Basic preprocess"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>pclass</th>\n",
       "      <th>name</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>sibsp</th>\n",
       "      <th>parch</th>\n",
       "      <th>ticket</th>\n",
       "      <th>fare</th>\n",
       "      <th>cabin</th>\n",
       "      <th>embarked</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>passengerid</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>Oconnor, Frankie</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>209245</td>\n",
       "      <td>27.14</td>\n",
       "      <td>C12239</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3</td>\n",
       "      <td>Bryan, Drew</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>27323</td>\n",
       "      <td>13.35</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>Owens, Kenneth</td>\n",
       "      <td>male</td>\n",
       "      <td>0.33</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>CA 457703</td>\n",
       "      <td>71.29</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>Kramer, James</td>\n",
       "      <td>male</td>\n",
       "      <td>19.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>A. 10866</td>\n",
       "      <td>13.04</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>Bond, Michael</td>\n",
       "      <td>male</td>\n",
       "      <td>25.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>427635</td>\n",
       "      <td>7.76</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             pclass              name   sex    age  sibsp  parch     ticket  \\\n",
       "passengerid                                                                   \n",
       "0                 1  Oconnor, Frankie  male    NaN      2      0     209245   \n",
       "1                 3       Bryan, Drew  male    NaN      0      0      27323   \n",
       "2                 3    Owens, Kenneth  male   0.33      1      2  CA 457703   \n",
       "3                 3     Kramer, James  male  19.00      0      0   A. 10866   \n",
       "4                 3     Bond, Michael  male  25.00      0      0     427635   \n",
       "\n",
       "              fare   cabin embarked  \n",
       "passengerid                          \n",
       "0            27.14  C12239        S  \n",
       "1            13.35     NaN        S  \n",
       "2            71.29     NaN        S  \n",
       "3            13.04     NaN        S  \n",
       "4             7.76     NaN        S  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train = df_train[[col for col in df_train.columns if col != 'Survived']]\n",
    "X_train.columns = [c.lower() for c in X_train.columns]\n",
    "X_train.set_index('passengerid', inplace=True)\n",
    "X_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "passengerid\n",
       "0    1\n",
       "1    0\n",
       "2    0\n",
       "3    0\n",
       "4    1\n",
       "Name: survived, dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train = df_train[['PassengerId','Survived']]\n",
    "y_train.columns = [c.lower() for c in y_train.columns]\n",
    "y_train.set_index('passengerid', inplace=True)\n",
    "y_train = y_train['survived']\n",
    "y_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>pclass</th>\n",
       "      <th>name</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>sibsp</th>\n",
       "      <th>parch</th>\n",
       "      <th>ticket</th>\n",
       "      <th>fare</th>\n",
       "      <th>cabin</th>\n",
       "      <th>embarked</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>passengerid</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>100000</th>\n",
       "      <td>3</td>\n",
       "      <td>Holliday, Daniel</td>\n",
       "      <td>male</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>24745</td>\n",
       "      <td>63.01</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100001</th>\n",
       "      <td>3</td>\n",
       "      <td>Nguyen, Lorraine</td>\n",
       "      <td>female</td>\n",
       "      <td>53.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13264</td>\n",
       "      <td>5.81</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100002</th>\n",
       "      <td>1</td>\n",
       "      <td>Harris, Heather</td>\n",
       "      <td>female</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>25990</td>\n",
       "      <td>38.91</td>\n",
       "      <td>B15315</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100003</th>\n",
       "      <td>2</td>\n",
       "      <td>Larsen, Eric</td>\n",
       "      <td>male</td>\n",
       "      <td>25.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>314011</td>\n",
       "      <td>12.93</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100004</th>\n",
       "      <td>1</td>\n",
       "      <td>Cleary, Sarah</td>\n",
       "      <td>female</td>\n",
       "      <td>17.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>26203</td>\n",
       "      <td>26.89</td>\n",
       "      <td>B22515</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             pclass              name     sex   age  sibsp  parch  ticket  \\\n",
       "passengerid                                                                 \n",
       "100000            3  Holliday, Daniel    male  19.0      0      0   24745   \n",
       "100001            3  Nguyen, Lorraine  female  53.0      0      0   13264   \n",
       "100002            1   Harris, Heather  female  19.0      0      0   25990   \n",
       "100003            2      Larsen, Eric    male  25.0      0      0  314011   \n",
       "100004            1     Cleary, Sarah  female  17.0      0      2   26203   \n",
       "\n",
       "              fare   cabin embarked  \n",
       "passengerid                          \n",
       "100000       63.01     NaN        S  \n",
       "100001        5.81     NaN        S  \n",
       "100002       38.91  B15315        C  \n",
       "100003       12.93     NaN        S  \n",
       "100004       26.89  B22515        C  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test = df_test[[col for col in df_test.columns if col != 'Survived']]\n",
    "X_test.columns = [c.lower() for c in X_test.columns]\n",
    "X_test.set_index('passengerid', inplace=True)\n",
    "X_test.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## pipeline simple"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "text_features = ['name', 'ticket']\n",
    "cat_features = ['sex', 'embarked', 'cabin']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from catboost import CatBoostClassifier\n",
    "from sklearn.naive_bayes import GaussianNB"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "original_features = X_train.columns.to_list()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def transform_name(df_base, col):\n",
    "    df = df_base[col].apply(lambda x: x.split(',')).to_frame('list_name')\n",
    "    df['surname'] = df['list_name'].apply(lambda x: x[0])\n",
    "    df['forename'] = df['list_name'].apply(lambda x: x[1])\n",
    "    return df[['surname', 'forename']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_tck_str(tck):\n",
    "    try:\n",
    "        x = tck.split()\n",
    "        try: \n",
    "            a = int(x[0])\n",
    "            return None\n",
    "        except:\n",
    "            return x[0]\n",
    "    except:\n",
    "        return None\n",
    "    \n",
    "def get_tck_num(tck):\n",
    "    try:\n",
    "        x = tck.split()\n",
    "        try:\n",
    "            a = int(x[0])\n",
    "            return a\n",
    "        except:\n",
    "            return None\n",
    "    except:\n",
    "        return None\n",
    "\n",
    "def transform_ticket(df_base, col):\n",
    "    df_ticket = df_base[col].to_frame('ticket')\n",
    "    df_ticket['ticket_str'] = df_base[col].apply(get_tck_str)\n",
    "    df_ticket['ticket_num'] = df_base[col].apply(get_tck_num)\n",
    "    return df_ticket[['ticket_str', 'ticket_num']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def transform_cabin(df_base, col):\n",
    "    df_cabin = df_base[col].to_frame('cabin')\n",
    "    df_cabin_clean = df_cabin[~df_cabin['cabin'].isna()].copy()\n",
    "    df_cabin_clean['cabin_str'] = df_cabin_clean['cabin'].apply(lambda x: x[0])\n",
    "    df_cabin_clean['cabin_num'] = df_cabin_clean['cabin'].apply(lambda x: x[1:]).astype(int)\n",
    "    return df_cabin.join(df_cabin_clean[['cabin_str', 'cabin_num']])[['cabin_str', 'cabin_num']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.base import TransformerMixin, BaseEstimator\n",
    "class PassNameTransformer(BaseEstimator, TransformerMixin):\n",
    "    def fit(self, X, y=None):\n",
    "        return self\n",
    "\n",
    "    def transform(self, X):\n",
    "        self.X = transform_name(X, 'name')\n",
    "        return self.X\n",
    "\n",
    "    def get_feature_names(self):\n",
    "        return self.X.columns.tolist()\n",
    "\n",
    "class TicketTransformer(BaseEstimator, TransformerMixin):\n",
    "    def fit(self, X, y=None):\n",
    "        return self\n",
    "\n",
    "    def transform(self, X):\n",
    "        self.X = transform_ticket(X, 'ticket')\n",
    "        return self.X\n",
    "\n",
    "    def get_feature_names(self):\n",
    "        return self.X.columns.tolist()\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def lower_values(df):\n",
    "    o_features = df.dtypes[df.dtypes=='object'].index.to_list()\n",
    "    df_copy = df.copy()\n",
    "    for c in o_features:\n",
    "        df_copy[c] = df_copy[c].apply(lambda x: x.lower() if type(x)==str else None)\n",
    "    return df_copy\n",
    "class LowerTransformer(BaseEstimator, TransformerMixin):\n",
    "    def fit(self, X, y=None):\n",
    "        return self\n",
    "    def transform(self,X):\n",
    "        return lower_values(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Programming Skills Supplement"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This section shows my proficiency in programming using Python, some of the demonstrated topics are:\n",
    "> - Loop and loop comprehension\n",
    "> - Object-oriented programming (OOP): class, object, custom method, inheritance, etc.\n",
    "> - Using numpy and pandas to manipulate arrays/dataframes\n",
    "> - Using scikit-learn's transformers and pipeline to transform data\n",
    "\n",
    "Note: Due to the length limit, some classes were imported before this section. You can find the full notebook with [this link](https://github.com/trantrikien239/pandas-sklearn-toolkit/blob/main/notebooks/pipeline_titanic.ipynb).\n",
    "In addition to the code provided here, please also take a look at some of my other works:\n",
    "- Blogpost on [applying sequence approach to grocery transactiond data](https://master-data.science/2021/11/30/Applying-sequence-classification-to-grocery-data-with-a-two-step-approach.html)\n",
    "- Github project where I created a [Sudoku solver from scratch](https://github.com/trantrikien239/sudoku)\n",
    "- Medium article on [team work in data science](https://medium.com/@kien239/how-to-make-delegation-work-in-a-data-science-team-beb5e8bd5901)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.pipeline import FeatureUnion\n",
    "from sklearn.base import TransformerMixin, BaseEstimator\n",
    "from tqdm.auto import tqdm\n",
    "from sklearn import set_config; set_config(display='diagram')\n",
    "\n",
    "class featureUnion(FeatureUnion):\n",
    "    def _hstack(self, Xs):\n",
    "        cols = [X.columns.tolist() for X in Xs]\n",
    "        dtypes = []\n",
    "        for X in Xs:\n",
    "            dtypes.append([str(X[col].dtype) for col in X])\n",
    "        cols = np.hstack(cols)\n",
    "        dtypes = np.hstack(dtypes)\n",
    "        data = pd.DataFrame(super()._hstack(Xs), columns = cols)\n",
    "        print('====Converting columns types====')\n",
    "        for col, dtype in tqdm(zip(cols, dtypes)):\n",
    "            data[col] = data[col].astype(dtype)\n",
    "        return data\n",
    "\n",
    "def transform_cabin(df_base, col):\n",
    "    df_cabin = df_base[col].to_frame('cabin')\n",
    "    df_cabin_clean = df_cabin[~df_cabin['cabin'].isna()].copy()\n",
    "    df_cabin_clean['cabin_str'] = df_cabin_clean['cabin'].apply(lambda x: x[0])\n",
    "    df_cabin_clean['cabin_num'] = df_cabin_clean['cabin'].apply(lambda x: x[1:]).astype(int)\n",
    "    return df_cabin.join(df_cabin_clean[['cabin_str', 'cabin_num']])[['cabin_str', 'cabin_num']]\n",
    "class CabinTransformer(BaseEstimator, TransformerMixin):\n",
    "    def fit(self, X, y=None):\n",
    "        return self\n",
    "    def transform(self, X):\n",
    "        self.X = transform_cabin(X, 'cabin')\n",
    "        return self.X\n",
    "    def get_feature_names(self):\n",
    "        return self.X.columns.tolist()\n",
    "\n",
    "class ColumnSelectTransformer(BaseEstimator, TransformerMixin):\n",
    "    def __init__(self, columns=None):\n",
    "        super().__init__()\n",
    "        self.columns=columns\n",
    "    def fit(self, X, y=None):\n",
    "        return self\n",
    "    def transform(self, X):\n",
    "        return X[self.columns]\n",
    "\n",
    "def lower_values(df):\n",
    "    o_features = df.dtypes[df.dtypes=='object'].index.to_list()\n",
    "    df_copy = df.copy()\n",
    "    for c in o_features:\n",
    "        df_copy[c] = df_copy[c].apply(lambda x: x.lower() if type(x)==str else None)\n",
    "    return df_copy\n",
    "class LowerTransformer(BaseEstimator, TransformerMixin):\n",
    "    def fit(self, X, y=None):\n",
    "        return self\n",
    "    def transform(self,X):\n",
    "        return lower_values(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "col_processor = featureUnion(transformer_list=[\n",
    "    ('pass_name', PassNameTransformer()),\n",
    "    ('ticket', TicketTransformer()),\n",
    "    ('cabin', CabinTransformer()),\n",
    "    ('others', ColumnSelectTransformer(columns=[c for c in original_features if c not in ('name', 'ticket', 'cabin')]))\n",
    "]\n",
    ")\n",
    "pl = Pipeline(steps=[\n",
    "    ('raw_data_processor', col_processor),\n",
    "    ('lower_text_values', LowerTransformer())\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 {color: black;background-color: white;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 pre{padding: 0;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-toggleable {background-color: white;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-estimator:hover {background-color: #d4ebff;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-item {z-index: 1;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-parallel::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-parallel-item:only-child::after {width: 0;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;position: relative;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-label-container {position: relative;z-index: 2;text-align: center;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-ecaee8db-163c-41d3-a322-a6f8fc1acbf0\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                 featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                 PassNameTransformer()),\n",
       "                                                (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                                                (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                                                (&#x27;others&#x27;,\n",
       "                                                 ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                  &#x27;sex&#x27;,\n",
       "                                                                                  &#x27;age&#x27;,\n",
       "                                                                                  &#x27;sibsp&#x27;,\n",
       "                                                                                  &#x27;parch&#x27;,\n",
       "                                                                                  &#x27;fare&#x27;,\n",
       "                                                                                  &#x27;embarked&#x27;]))])),\n",
       "                (&#x27;lower_text_values&#x27;, LowerTransformer())])</pre><b>Please rerun this cell to show the HTML repr or trust the notebook.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"f9eecb4b-875f-4629-adc8-2dfdca6ad6f4\" type=\"checkbox\" ><label for=\"f9eecb4b-875f-4629-adc8-2dfdca6ad6f4\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                 featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                 PassNameTransformer()),\n",
       "                                                (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                                                (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                                                (&#x27;others&#x27;,\n",
       "                                                 ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                  &#x27;sex&#x27;,\n",
       "                                                                                  &#x27;age&#x27;,\n",
       "                                                                                  &#x27;sibsp&#x27;,\n",
       "                                                                                  &#x27;parch&#x27;,\n",
       "                                                                                  &#x27;fare&#x27;,\n",
       "                                                                                  &#x27;embarked&#x27;]))])),\n",
       "                (&#x27;lower_text_values&#x27;, LowerTransformer())])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"16ff1d66-1688-438f-a798-66755e551557\" type=\"checkbox\" ><label for=\"16ff1d66-1688-438f-a798-66755e551557\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">raw_data_processor: featureUnion</label><div class=\"sk-toggleable__content\"><pre>featureUnion(transformer_list=[(&#x27;pass_name&#x27;, PassNameTransformer()),\n",
       "                               (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                               (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                               (&#x27;others&#x27;,\n",
       "                                ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                 &#x27;sex&#x27;, &#x27;age&#x27;,\n",
       "                                                                 &#x27;sibsp&#x27;,\n",
       "                                                                 &#x27;parch&#x27;,\n",
       "                                                                 &#x27;fare&#x27;,\n",
       "                                                                 &#x27;embarked&#x27;]))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>pass_name</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"fdd3d1bb-ff32-49d1-966a-b57123da75d4\" type=\"checkbox\" ><label for=\"fdd3d1bb-ff32-49d1-966a-b57123da75d4\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">PassNameTransformer</label><div class=\"sk-toggleable__content\"><pre>PassNameTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>ticket</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"2e2f9d84-65e9-4b26-a3a7-f51d4c15329f\" type=\"checkbox\" ><label for=\"2e2f9d84-65e9-4b26-a3a7-f51d4c15329f\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">TicketTransformer</label><div class=\"sk-toggleable__content\"><pre>TicketTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>cabin</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"9289c8f9-5e29-41f2-a0d5-83462adf7389\" type=\"checkbox\" ><label for=\"9289c8f9-5e29-41f2-a0d5-83462adf7389\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CabinTransformer</label><div class=\"sk-toggleable__content\"><pre>CabinTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>others</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"5aedfa5a-91b1-4fd9-9be4-3c01e50966e6\" type=\"checkbox\" ><label for=\"5aedfa5a-91b1-4fd9-9be4-3c01e50966e6\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ColumnSelectTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnSelectTransformer(columns=[&#x27;pclass&#x27;, &#x27;sex&#x27;, &#x27;age&#x27;, &#x27;sibsp&#x27;, &#x27;parch&#x27;,\n",
       "                                 &#x27;fare&#x27;, &#x27;embarked&#x27;])</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"9e8b2439-8de9-49be-9651-af960ae7c910\" type=\"checkbox\" ><label for=\"9e8b2439-8de9-49be-9651-af960ae7c910\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">LowerTransformer</label><div class=\"sk-toggleable__content\"><pre>LowerTransformer()</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('raw_data_processor',\n",
       "                 featureUnion(transformer_list=[('pass_name',\n",
       "                                                 PassNameTransformer()),\n",
       "                                                ('ticket', TicketTransformer()),\n",
       "                                                ('cabin', CabinTransformer()),\n",
       "                                                ('others',\n",
       "                                                 ColumnSelectTransformer(columns=['pclass',\n",
       "                                                                                  'sex',\n",
       "                                                                                  'age',\n",
       "                                                                                  'sibsp',\n",
       "                                                                                  'parch',\n",
       "                                                                                  'fare',\n",
       "                                                                                  'embarked']))])),\n",
       "                ('lower_text_values', LowerTransformer())])"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>pclass</th>\n",
       "      <th>name</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>sibsp</th>\n",
       "      <th>parch</th>\n",
       "      <th>ticket</th>\n",
       "      <th>fare</th>\n",
       "      <th>cabin</th>\n",
       "      <th>embarked</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>passengerid</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>Oconnor, Frankie</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>209245</td>\n",
       "      <td>27.14</td>\n",
       "      <td>C12239</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3</td>\n",
       "      <td>Bryan, Drew</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>27323</td>\n",
       "      <td>13.35</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>Owens, Kenneth</td>\n",
       "      <td>male</td>\n",
       "      <td>0.33</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>CA 457703</td>\n",
       "      <td>71.29</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>Kramer, James</td>\n",
       "      <td>male</td>\n",
       "      <td>19.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>A. 10866</td>\n",
       "      <td>13.04</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>Bond, Michael</td>\n",
       "      <td>male</td>\n",
       "      <td>25.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>427635</td>\n",
       "      <td>7.76</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             pclass              name   sex    age  sibsp  parch     ticket  \\\n",
       "passengerid                                                                   \n",
       "0                 1  Oconnor, Frankie  male    NaN      2      0     209245   \n",
       "1                 3       Bryan, Drew  male    NaN      0      0      27323   \n",
       "2                 3    Owens, Kenneth  male   0.33      1      2  CA 457703   \n",
       "3                 3     Kramer, James  male  19.00      0      0   A. 10866   \n",
       "4                 3     Bond, Michael  male  25.00      0      0     427635   \n",
       "\n",
       "              fare   cabin embarked  \n",
       "passengerid                          \n",
       "0            27.14  C12239        S  \n",
       "1            13.35     NaN        S  \n",
       "2            71.29     NaN        S  \n",
       "3            13.04     NaN        S  \n",
       "4             7.76     NaN        S  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4aa0423d927443979eace4e5053fb020",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>surname</th>\n",
       "      <th>forename</th>\n",
       "      <th>ticket_str</th>\n",
       "      <th>ticket_num</th>\n",
       "      <th>cabin_str</th>\n",
       "      <th>cabin_num</th>\n",
       "      <th>pclass</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>sibsp</th>\n",
       "      <th>parch</th>\n",
       "      <th>fare</th>\n",
       "      <th>embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>oconnor</td>\n",
       "      <td>frankie</td>\n",
       "      <td>None</td>\n",
       "      <td>209245.0</td>\n",
       "      <td>c</td>\n",
       "      <td>12239.0</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>27.14</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>bryan</td>\n",
       "      <td>drew</td>\n",
       "      <td>None</td>\n",
       "      <td>27323.0</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13.35</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>owens</td>\n",
       "      <td>kenneth</td>\n",
       "      <td>ca</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>0.33</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>71.29</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>kramer</td>\n",
       "      <td>james</td>\n",
       "      <td>a.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>19.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13.04</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>bond</td>\n",
       "      <td>michael</td>\n",
       "      <td>None</td>\n",
       "      <td>427635.0</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>25.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.76</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   surname  forename ticket_str  ticket_num cabin_str  cabin_num  pclass  \\\n",
       "0  oconnor   frankie       None    209245.0         c    12239.0       1   \n",
       "1    bryan      drew       None     27323.0      None        NaN       3   \n",
       "2    owens   kenneth         ca         NaN      None        NaN       3   \n",
       "3   kramer     james         a.         NaN      None        NaN       3   \n",
       "4     bond   michael       None    427635.0      None        NaN       3   \n",
       "\n",
       "    sex    age  sibsp  parch   fare embarked  \n",
       "0  male    NaN      2      0  27.14        s  \n",
       "1  male    NaN      0      0  13.35        s  \n",
       "2  male   0.33      1      2  71.29        s  \n",
       "3  male  19.00      0      0  13.04        s  \n",
       "4  male  25.00      0      0   7.76        s  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2 = pl.fit_transform(X_train)\n",
    "df2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c87324502cbe45fa972521d95ee6f619",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>surname</th>\n",
       "      <th>forename</th>\n",
       "      <th>ticket_str</th>\n",
       "      <th>ticket_num</th>\n",
       "      <th>cabin_str</th>\n",
       "      <th>cabin_num</th>\n",
       "      <th>pclass</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>sibsp</th>\n",
       "      <th>parch</th>\n",
       "      <th>fare</th>\n",
       "      <th>embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>oconnor</td>\n",
       "      <td>frankie</td>\n",
       "      <td>None</td>\n",
       "      <td>209245.0</td>\n",
       "      <td>c</td>\n",
       "      <td>12239.0</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>27.14</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>bryan</td>\n",
       "      <td>drew</td>\n",
       "      <td>None</td>\n",
       "      <td>27323.0</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13.35</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>owens</td>\n",
       "      <td>kenneth</td>\n",
       "      <td>ca</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>0.33</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>71.29</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>kramer</td>\n",
       "      <td>james</td>\n",
       "      <td>a.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>19.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13.04</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>bond</td>\n",
       "      <td>michael</td>\n",
       "      <td>None</td>\n",
       "      <td>427635.0</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>25.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.76</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   surname  forename ticket_str  ticket_num cabin_str  cabin_num  pclass  \\\n",
       "0  oconnor   frankie       None    209245.0         c    12239.0       1   \n",
       "1    bryan      drew       None     27323.0      None        NaN       3   \n",
       "2    owens   kenneth         ca         NaN      None        NaN       3   \n",
       "3   kramer     james         a.         NaN      None        NaN       3   \n",
       "4     bond   michael       None    427635.0      None        NaN       3   \n",
       "\n",
       "    sex    age  sibsp  parch   fare embarked  \n",
       "0  male    NaN      2      0  27.14        s  \n",
       "1  male    NaN      0      0  13.35        s  \n",
       "2  male   0.33      1      2  71.29        s  \n",
       "3  male  19.00      0      0  13.04        s  \n",
       "4  male  25.00      0      0   7.76        s  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2 = pl.fit_transform(X_train)\n",
    "df2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.impute import SimpleImputer\n",
    "\n",
    "from sklearn.compose import ColumnTransformer, make_column_selector\n",
    "\n",
    "class columnTransformer(ColumnTransformer):\n",
    "    def _hstack(self, Xs):\n",
    "        cols = [X.columns.tolist() for X in Xs]\n",
    "        dtypes = []\n",
    "        print(cols)\n",
    "        print([X.shape for X in Xs])\n",
    "        for X in Xs:\n",
    "            dtypes.append([str(X[col].dtype) for col in X])\n",
    "        cols = np.hstack(cols)\n",
    "        dtypes = np.hstack(dtypes)\n",
    "        data = pd.DataFrame(super()._hstack(Xs), columns = cols)\n",
    "        print('====Converting columns types====')\n",
    "        for col, dtype in tqdm(zip(cols, dtypes)):\n",
    "            data[col] = data[col].astype(dtype)\n",
    "        return data\n",
    "\n",
    "class simpleImputer(SimpleImputer):\n",
    "    def fit(self, X, y=None):\n",
    "        self._cols = X.columns.tolist()\n",
    "        self._dtypes = [str(X[col].dtype) for col in X.columns]\n",
    "        super().fit(X, y)\n",
    "        return self\n",
    "        \n",
    "    def transform(self, X):\n",
    "        X_ = super().transform(X)\n",
    "        data = pd.DataFrame(X_, columns = self._cols)\n",
    "        for col, dtype in tqdm(zip(self._cols, self._dtypes)):\n",
    "            data[col] = data[col].astype(dtype)\n",
    "        return data\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "col_handle_na = columnTransformer(\n",
    "    transformers=[\n",
    "        ('text_features', simpleImputer(missing_values=None, strategy='constant', fill_value='unk'), make_column_selector(dtype_include=['object'])),\n",
    "        ('float_features', simpleImputer(strategy='median'), make_column_selector(dtype_include=['float64'])),\n",
    "        ('count_features', simpleImputer(strategy='most_frequent'), make_column_selector(dtype_include=['int64']))\n",
    "    ],\n",
    "    remainder='passthrough'\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "pl_simple = Pipeline(steps=[\n",
    "    ('pl', pl),\n",
    "    ('null_handling', col_handle_na)\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 {color: black;background-color: white;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 pre{padding: 0;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-toggleable {background-color: white;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-estimator:hover {background-color: #d4ebff;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-item {z-index: 1;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-parallel::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-parallel-item:only-child::after {width: 0;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;position: relative;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-label-container {position: relative;z-index: 2;text-align: center;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-6a9345d5-23e2-47ff-94cc-5967bcbcbdf0\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;pl&#x27;,\n",
       "                 Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                                  featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                                  PassNameTransformer()),\n",
       "                                                                 (&#x27;ticket&#x27;,\n",
       "                                                                  TicketTransformer()),\n",
       "                                                                 (&#x27;cabin&#x27;,\n",
       "                                                                  CabinTransformer()),\n",
       "                                                                 (&#x27;others&#x27;,\n",
       "                                                                  ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                                   &#x27;sex&#x27;,\n",
       "                                                                                                   &#x27;age&#x27;,\n",
       "                                                                                                   &#x27;sibsp&#x27;,\n",
       "                                                                                                   &#x27;parch&#x27;,\n",
       "                                                                                                   &#x27;fare&#x27;,\n",
       "                                                                                                   &#x27;embarked&#x27;]))])),\n",
       "                                 (&#x27;lower_text_values&#x27;, LowerTransformer())])),\n",
       "                (&#x27;null_handling...\n",
       "                                                  &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9f10&gt;),\n",
       "                                                 (&#x27;float_features&#x27;,\n",
       "                                                  simpleImputer(strategy=&#x27;median&#x27;),\n",
       "                                                  &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9850&gt;),\n",
       "                                                 (&#x27;count_features&#x27;,\n",
       "                                                  simpleImputer(strategy=&#x27;most_frequent&#x27;),\n",
       "                                                  &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070&gt;)]))])</pre><b>Please rerun this cell to show the HTML repr or trust the notebook.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"3efaba20-84bd-4ca2-8079-9f98f6247559\" type=\"checkbox\" ><label for=\"3efaba20-84bd-4ca2-8079-9f98f6247559\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;pl&#x27;,\n",
       "                 Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                                  featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                                  PassNameTransformer()),\n",
       "                                                                 (&#x27;ticket&#x27;,\n",
       "                                                                  TicketTransformer()),\n",
       "                                                                 (&#x27;cabin&#x27;,\n",
       "                                                                  CabinTransformer()),\n",
       "                                                                 (&#x27;others&#x27;,\n",
       "                                                                  ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                                   &#x27;sex&#x27;,\n",
       "                                                                                                   &#x27;age&#x27;,\n",
       "                                                                                                   &#x27;sibsp&#x27;,\n",
       "                                                                                                   &#x27;parch&#x27;,\n",
       "                                                                                                   &#x27;fare&#x27;,\n",
       "                                                                                                   &#x27;embarked&#x27;]))])),\n",
       "                                 (&#x27;lower_text_values&#x27;, LowerTransformer())])),\n",
       "                (&#x27;null_handling...\n",
       "                                                  &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9f10&gt;),\n",
       "                                                 (&#x27;float_features&#x27;,\n",
       "                                                  simpleImputer(strategy=&#x27;median&#x27;),\n",
       "                                                  &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9850&gt;),\n",
       "                                                 (&#x27;count_features&#x27;,\n",
       "                                                  simpleImputer(strategy=&#x27;most_frequent&#x27;),\n",
       "                                                  &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070&gt;)]))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"aea81652-a06e-4b56-824e-7479a5aaa866\" type=\"checkbox\" ><label for=\"aea81652-a06e-4b56-824e-7479a5aaa866\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">pl: Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                 featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                 PassNameTransformer()),\n",
       "                                                (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                                                (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                                                (&#x27;others&#x27;,\n",
       "                                                 ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                  &#x27;sex&#x27;,\n",
       "                                                                                  &#x27;age&#x27;,\n",
       "                                                                                  &#x27;sibsp&#x27;,\n",
       "                                                                                  &#x27;parch&#x27;,\n",
       "                                                                                  &#x27;fare&#x27;,\n",
       "                                                                                  &#x27;embarked&#x27;]))])),\n",
       "                (&#x27;lower_text_values&#x27;, LowerTransformer())])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"72becb89-d04e-41da-ae5a-2e3f70b8da4d\" type=\"checkbox\" ><label for=\"72becb89-d04e-41da-ae5a-2e3f70b8da4d\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">raw_data_processor: featureUnion</label><div class=\"sk-toggleable__content\"><pre>featureUnion(transformer_list=[(&#x27;pass_name&#x27;, PassNameTransformer()),\n",
       "                               (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                               (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                               (&#x27;others&#x27;,\n",
       "                                ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                 &#x27;sex&#x27;, &#x27;age&#x27;,\n",
       "                                                                 &#x27;sibsp&#x27;,\n",
       "                                                                 &#x27;parch&#x27;,\n",
       "                                                                 &#x27;fare&#x27;,\n",
       "                                                                 &#x27;embarked&#x27;]))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>pass_name</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"77445abb-b965-4ce4-9871-63e4e1e1efc6\" type=\"checkbox\" ><label for=\"77445abb-b965-4ce4-9871-63e4e1e1efc6\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">PassNameTransformer</label><div class=\"sk-toggleable__content\"><pre>PassNameTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>ticket</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"2441371a-b089-4d44-af5e-564464516231\" type=\"checkbox\" ><label for=\"2441371a-b089-4d44-af5e-564464516231\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">TicketTransformer</label><div class=\"sk-toggleable__content\"><pre>TicketTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>cabin</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"f4cbe5c3-7a28-4e60-bc99-be7e85cb065e\" type=\"checkbox\" ><label for=\"f4cbe5c3-7a28-4e60-bc99-be7e85cb065e\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CabinTransformer</label><div class=\"sk-toggleable__content\"><pre>CabinTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>others</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"9b45fb87-ebfb-4ee0-85ba-53e5f06bfeba\" type=\"checkbox\" ><label for=\"9b45fb87-ebfb-4ee0-85ba-53e5f06bfeba\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ColumnSelectTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnSelectTransformer(columns=[&#x27;pclass&#x27;, &#x27;sex&#x27;, &#x27;age&#x27;, &#x27;sibsp&#x27;, &#x27;parch&#x27;,\n",
       "                                 &#x27;fare&#x27;, &#x27;embarked&#x27;])</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"c2353801-9a14-4ab6-8898-0542dc7259d1\" type=\"checkbox\" ><label for=\"c2353801-9a14-4ab6-8898-0542dc7259d1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">LowerTransformer</label><div class=\"sk-toggleable__content\"><pre>LowerTransformer()</pre></div></div></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"e53775da-d0ce-4b2a-985e-d8d16ea441f1\" type=\"checkbox\" ><label for=\"e53775da-d0ce-4b2a-985e-d8d16ea441f1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">null_handling: columnTransformer</label><div class=\"sk-toggleable__content\"><pre>columnTransformer(remainder=&#x27;passthrough&#x27;,\n",
       "                  transformers=[(&#x27;text_features&#x27;,\n",
       "                                 simpleImputer(fill_value=&#x27;unk&#x27;,\n",
       "                                               missing_values=None,\n",
       "                                               strategy=&#x27;constant&#x27;),\n",
       "                                 &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9f10&gt;),\n",
       "                                (&#x27;float_features&#x27;,\n",
       "                                 simpleImputer(strategy=&#x27;median&#x27;),\n",
       "                                 &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9850&gt;),\n",
       "                                (&#x27;count_features&#x27;,\n",
       "                                 simpleImputer(strategy=&#x27;most_frequent&#x27;),\n",
       "                                 &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070&gt;)])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"27565445-e15b-4d4b-9159-9cb5f6560d4a\" type=\"checkbox\" ><label for=\"27565445-e15b-4d4b-9159-9cb5f6560d4a\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">text_features</label><div class=\"sk-toggleable__content\"><pre>&lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9f10&gt;</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"c306007e-8955-44fb-953a-acbec41100bb\" type=\"checkbox\" ><label for=\"c306007e-8955-44fb-953a-acbec41100bb\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">simpleImputer</label><div class=\"sk-toggleable__content\"><pre>simpleImputer(fill_value=&#x27;unk&#x27;, missing_values=None, strategy=&#x27;constant&#x27;)</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"b41bdf9e-d959-4d8f-818f-ab7465dc9995\" type=\"checkbox\" ><label for=\"b41bdf9e-d959-4d8f-818f-ab7465dc9995\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">float_features</label><div class=\"sk-toggleable__content\"><pre>&lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9850&gt;</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"d3ce5d47-5643-409d-9a01-a627be4a5f5a\" type=\"checkbox\" ><label for=\"d3ce5d47-5643-409d-9a01-a627be4a5f5a\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">simpleImputer</label><div class=\"sk-toggleable__content\"><pre>simpleImputer(strategy=&#x27;median&#x27;)</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"35c620f7-e45b-4f11-aa4d-acb60f42c04c\" type=\"checkbox\" ><label for=\"35c620f7-e45b-4f11-aa4d-acb60f42c04c\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">count_features</label><div class=\"sk-toggleable__content\"><pre>&lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070&gt;</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"117126ab-3642-4e1e-ac18-f5c5137c84d9\" type=\"checkbox\" ><label for=\"117126ab-3642-4e1e-ac18-f5c5137c84d9\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">simpleImputer</label><div class=\"sk-toggleable__content\"><pre>simpleImputer(strategy=&#x27;most_frequent&#x27;)</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"7f8f048d-53c6-4b39-a73f-5c06f493c3e0\" type=\"checkbox\" ><label for=\"7f8f048d-53c6-4b39-a73f-5c06f493c3e0\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">remainder</label><div class=\"sk-toggleable__content\"><pre></pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"e03bf769-684a-4630-8872-5a08aa0e4fa3\" type=\"checkbox\" ><label for=\"e03bf769-684a-4630-8872-5a08aa0e4fa3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">passthrough</label><div class=\"sk-toggleable__content\"><pre>passthrough</pre></div></div></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('pl',\n",
       "                 Pipeline(steps=[('raw_data_processor',\n",
       "                                  featureUnion(transformer_list=[('pass_name',\n",
       "                                                                  PassNameTransformer()),\n",
       "                                                                 ('ticket',\n",
       "                                                                  TicketTransformer()),\n",
       "                                                                 ('cabin',\n",
       "                                                                  CabinTransformer()),\n",
       "                                                                 ('others',\n",
       "                                                                  ColumnSelectTransformer(columns=['pclass',\n",
       "                                                                                                   'sex',\n",
       "                                                                                                   'age',\n",
       "                                                                                                   'sibsp',\n",
       "                                                                                                   'parch',\n",
       "                                                                                                   'fare',\n",
       "                                                                                                   'embarked']))])),\n",
       "                                 ('lower_text_values', LowerTransformer())])),\n",
       "                ('null_handling...\n",
       "                                                  <sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9f10>),\n",
       "                                                 ('float_features',\n",
       "                                                  simpleImputer(strategy='median'),\n",
       "                                                  <sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9850>),\n",
       "                                                 ('count_features',\n",
       "                                                  simpleImputer(strategy='most_frequent'),\n",
       "                                                  <sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070>)]))])"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl_simple"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "95ed26a54830442ba6c084f4505eab78",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f2692d7a23d3461d871e4ac375b8497f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a9430b670d4e4ce39ddfaac382ad6725",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7239fb61c6e1426c95a45d4d443ec530",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[['surname', 'forename', 'ticket_str', 'cabin_str', 'sex', 'embarked'], ['ticket_num', 'cabin_num', 'age', 'fare'], ['pclass', 'sibsp', 'parch']]\n",
      "[(100000, 6), (100000, 4), (100000, 3)]\n",
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "91f9edda7941454aa9f5baed1caa80cc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df3 = pl_simple.fit_transform(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>surname</th>\n",
       "      <th>forename</th>\n",
       "      <th>ticket_str</th>\n",
       "      <th>cabin_str</th>\n",
       "      <th>sex</th>\n",
       "      <th>embarked</th>\n",
       "      <th>ticket_num</th>\n",
       "      <th>cabin_num</th>\n",
       "      <th>age</th>\n",
       "      <th>fare</th>\n",
       "      <th>pclass</th>\n",
       "      <th>sibsp</th>\n",
       "      <th>parch</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>oconnor</td>\n",
       "      <td>frankie</td>\n",
       "      <td>unk</td>\n",
       "      <td>c</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "      <td>209245.0</td>\n",
       "      <td>12239.0</td>\n",
       "      <td>39.00</td>\n",
       "      <td>27.14</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>bryan</td>\n",
       "      <td>drew</td>\n",
       "      <td>unk</td>\n",
       "      <td>unk</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "      <td>27323.0</td>\n",
       "      <td>12603.5</td>\n",
       "      <td>39.00</td>\n",
       "      <td>13.35</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>owens</td>\n",
       "      <td>kenneth</td>\n",
       "      <td>ca</td>\n",
       "      <td>unk</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "      <td>39185.5</td>\n",
       "      <td>12603.5</td>\n",
       "      <td>0.33</td>\n",
       "      <td>71.29</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>kramer</td>\n",
       "      <td>james</td>\n",
       "      <td>a.</td>\n",
       "      <td>unk</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "      <td>39185.5</td>\n",
       "      <td>12603.5</td>\n",
       "      <td>19.00</td>\n",
       "      <td>13.04</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>bond</td>\n",
       "      <td>michael</td>\n",
       "      <td>unk</td>\n",
       "      <td>unk</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "      <td>427635.0</td>\n",
       "      <td>12603.5</td>\n",
       "      <td>25.00</td>\n",
       "      <td>7.76</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99995</th>\n",
       "      <td>bell</td>\n",
       "      <td>adele</td>\n",
       "      <td>pc</td>\n",
       "      <td>d</td>\n",
       "      <td>female</td>\n",
       "      <td>c</td>\n",
       "      <td>39185.5</td>\n",
       "      <td>17243.0</td>\n",
       "      <td>62.00</td>\n",
       "      <td>14.86</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99996</th>\n",
       "      <td>brown</td>\n",
       "      <td>herman</td>\n",
       "      <td>unk</td>\n",
       "      <td>unk</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "      <td>13273.0</td>\n",
       "      <td>12603.5</td>\n",
       "      <td>66.00</td>\n",
       "      <td>11.15</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99997</th>\n",
       "      <td>childress</td>\n",
       "      <td>charles</td>\n",
       "      <td>unk</td>\n",
       "      <td>unk</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "      <td>39185.5</td>\n",
       "      <td>12603.5</td>\n",
       "      <td>37.00</td>\n",
       "      <td>9.95</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99998</th>\n",
       "      <td>caughlin</td>\n",
       "      <td>thomas</td>\n",
       "      <td>unk</td>\n",
       "      <td>unk</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "      <td>458654.0</td>\n",
       "      <td>12603.5</td>\n",
       "      <td>51.00</td>\n",
       "      <td>30.92</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99999</th>\n",
       "      <td>enciso</td>\n",
       "      <td>tyler</td>\n",
       "      <td>unk</td>\n",
       "      <td>unk</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "      <td>458074.0</td>\n",
       "      <td>12603.5</td>\n",
       "      <td>55.00</td>\n",
       "      <td>13.96</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100000 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         surname  forename ticket_str cabin_str     sex embarked  ticket_num  \\\n",
       "0        oconnor   frankie        unk         c    male        s    209245.0   \n",
       "1          bryan      drew        unk       unk    male        s     27323.0   \n",
       "2          owens   kenneth         ca       unk    male        s     39185.5   \n",
       "3         kramer     james         a.       unk    male        s     39185.5   \n",
       "4           bond   michael        unk       unk    male        s    427635.0   \n",
       "...          ...       ...        ...       ...     ...      ...         ...   \n",
       "99995       bell     adele         pc         d  female        c     39185.5   \n",
       "99996      brown    herman        unk       unk    male        s     13273.0   \n",
       "99997  childress   charles        unk       unk    male        s     39185.5   \n",
       "99998   caughlin    thomas        unk       unk    male        s    458654.0   \n",
       "99999     enciso     tyler        unk       unk    male        s    458074.0   \n",
       "\n",
       "       cabin_num    age   fare  pclass  sibsp  parch  \n",
       "0        12239.0  39.00  27.14       1      2      0  \n",
       "1        12603.5  39.00  13.35       3      0      0  \n",
       "2        12603.5   0.33  71.29       3      1      2  \n",
       "3        12603.5  19.00  13.04       3      0      0  \n",
       "4        12603.5  25.00   7.76       3      0      0  \n",
       "...          ...    ...    ...     ...    ...    ...  \n",
       "99995    17243.0  62.00  14.86       2      0      0  \n",
       "99996    12603.5  66.00  11.15       2      0      0  \n",
       "99997    12603.5  37.00   9.95       3      0      0  \n",
       "99998    12603.5  51.00  30.92       3      0      1  \n",
       "99999    12603.5  55.00  13.96       3      0      0  \n",
       "\n",
       "[100000 rows x 13 columns]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "class PipelineLogger(object):\n",
    "    def __init__(self):\n",
    "        pass\n",
    "        \n",
    "    def log_start(self):\n",
    "        self.start_time = time()\n",
    "        print(f'======== {self.__class__.__name__} - START ========')\n",
    "        return None\n",
    "        \n",
    "    def log_finish(self):\n",
    "        self.duration = time() - self.start_time\n",
    "        print(f'======== {self.__class__.__name__} - FINISH =======> Take: {self.duration:.6f}(s)')\n",
    "\n",
    "\n",
    "class ExperimentBase(BaseEstimator):\n",
    "    def evaluate(self, X_test, y_test):\n",
    "        print('Evaluating model')\n",
    "        print(classification_report(y_true=y_test, y_pred=self.predict(X_test)))\n",
    "        metrics = self.auc_report(X_test, y_test)\n",
    "        metrics['precision'], metrics['recall'], metrics['f1_score'], metrics['support'] = precision_recall_fscore_support(y_test, self.predict(X_test))\n",
    "        return metrics\n",
    "    \n",
    "    def auc_report(self, X, y_true):\n",
    "        classes = self.classes_\n",
    "        y_pred_classes = self.predict_proba(X)\n",
    "        n_classes = len(classes)\n",
    "\n",
    "        lw = 2\n",
    "        for i in range(len(classes)):\n",
    "            print(f\"\"\"{classes[i]}: {roc_auc_score(y_true=(y_true==classes[i]).astype(int), y_score=y_pred_classes[:,i])}\"\"\")\n",
    "\n",
    "        # Compute ROC curve and ROC area for each class\n",
    "        fpr = dict()\n",
    "        tpr = dict()\n",
    "        roc_auc = dict()\n",
    "\n",
    "        for i in range(n_classes):\n",
    "            fpr[i], tpr[i], _ = roc_curve(y_true=(y_true==classes[i]).astype(int), y_score=y_pred_classes[:,i])\n",
    "            roc_auc[i] = auc(fpr[i], tpr[i])\n",
    "\n",
    "        all_fpr = np.unique(np.concatenate([fpr[i] for i in range(len(classes))]))\n",
    "\n",
    "        # Then interpolate all ROC curves at this points\n",
    "        mean_tpr = np.zeros_like(all_fpr)\n",
    "        for i in range(len(classes)):\n",
    "            mean_tpr += interp(all_fpr, fpr[i], tpr[i])\n",
    "\n",
    "        # Finally average it and compute AUC\n",
    "        mean_tpr /= n_classes\n",
    "\n",
    "        fpr[\"macro\"] = all_fpr\n",
    "        tpr[\"macro\"] = mean_tpr\n",
    "        roc_auc[\"macro\"] = auc(fpr[\"macro\"], tpr[\"macro\"])\n",
    "\n",
    "        # Plot all ROC curves\n",
    "        plt.figure()\n",
    "\n",
    "        plt.plot(fpr[\"macro\"], tpr[\"macro\"],\n",
    "                 label='macro-average ROC curve (area = {0:0.2f})'\n",
    "                       ''.format(roc_auc[\"macro\"]),\n",
    "                 color='navy', linestyle=':', linewidth=4)\n",
    "\n",
    "        colors = cycle(['aqua', 'darkorange', 'cornflowerblue'])\n",
    "        for i, color in zip(range(n_classes), colors):\n",
    "            plt.plot(fpr[i], tpr[i], color=color, lw=lw,\n",
    "                     label='ROC curve of class {0} (area = {1:0.2f})'\n",
    "                     ''.format(classes[i], roc_auc[i]))\n",
    "\n",
    "        plt.plot([0, 1], [0, 1], 'k--', lw=lw)\n",
    "        plt.xlim([0.0, 1.0])\n",
    "        plt.ylim([0.0, 1.05])\n",
    "        plt.xlabel('False Positive Rate')\n",
    "        plt.ylabel('True Positive Rate')\n",
    "        plt.title('Some extension of Receiver operating characteristic to multi-class')\n",
    "        plt.legend(loc=\"lower right\")\n",
    "        plt.show()\n",
    "        metrics = {\n",
    "            'macro_auc': roc_auc[\"macro\"]\n",
    "        }\n",
    "        for i in range(n_classes):\n",
    "            metrics[f'auc_{classes[i]}'] = roc_auc[i]\n",
    "        return metrics\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "class CustomCatBoostClassifier(CatBoostClassifier, ExperimentBase, PipelineLogger):\n",
    "    def __init__(self, *args, **kwargs):\n",
    "        super().__init__(*args, **kwargs)\n",
    "        \n",
    "    def fit(self, X, y, self_evaluate=True, **kwargs):\n",
    "        self.log_start()\n",
    "        if self._init_params.get('cat_features') is not None:\n",
    "            cat_features_ = [c for c in self._init_params['cat_features'] if c in X.columns]\n",
    "            self._init_params['cat_features'] = cat_features_\n",
    "        else:\n",
    "            cat_features_ = None\n",
    "        if self._init_params.get('text_features') is not None:\n",
    "            text_features_ = [c for c in self._init_params['text_features'] if c in X.columns]\n",
    "            self._init_params['text_features'] = text_features_\n",
    "        else:\n",
    "            text_features_ = None\n",
    "        \n",
    "        X_t, X_e, y_t, y_e = self.train_eval_split(X, y, cat_features_, text_features_)\n",
    "        super().fit(X_t, y_t, eval_set=(X_e, y_e), cat_features=cat_features_, text_features=text_features_)\n",
    "        if self_evaluate:\n",
    "            _ = self.evaluate(X_e, y_e)\n",
    "        self.log_finish()\n",
    "        return self\n",
    "        \n",
    "    def train_eval_split(self, X, y, cat_features_, text_features_, eval_frac=0.1, add_na_Xy=False, na_label=0):\n",
    "        X_e = X.sample(frac=eval_frac, random_state=42)\n",
    "        y_e = y.loc[X_e.index]\n",
    "        X_t = X.drop(X_e.index)\n",
    "        y_t = y.loc[X_t.index]\n",
    "        if add_na_Xy:\n",
    "            X_t = pd.concat([X_t, pd.DataFrame([[np.nan] * X_t.shape[1]], columns=X_t.columns)], ignore_index=True)\n",
    "            y_t = pd.concat([y_t, pd.Series([na_label])], ignore_index=True)\n",
    "        if cat_features_ is not None:\n",
    "            cat_features_ = [c for c in cat_features_ if c in X.columns]\n",
    "            X_t[cat_features_] = X_t[cat_features_].fillna('unk')\n",
    "            X_e[cat_features_] = X_e[cat_features_].fillna('unk')\n",
    "        if text_features_ is not None:\n",
    "            text_features_ = [c for c in text_features_ if c in X.columns]\n",
    "            X_t[text_features_] = X_t[text_features_].fillna('unk')\n",
    "            X_e[text_features_] = X_e[text_features_].fillna('unk')\n",
    "        \n",
    "        return X_t, X_e, y_t, y_e\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>surname</th>\n",
       "      <th>forename</th>\n",
       "      <th>ticket_str</th>\n",
       "      <th>cabin_str</th>\n",
       "      <th>sex</th>\n",
       "      <th>embarked</th>\n",
       "      <th>ticket_num</th>\n",
       "      <th>cabin_num</th>\n",
       "      <th>age</th>\n",
       "      <th>fare</th>\n",
       "      <th>pclass</th>\n",
       "      <th>sibsp</th>\n",
       "      <th>parch</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>oconnor</td>\n",
       "      <td>frankie</td>\n",
       "      <td>unk</td>\n",
       "      <td>c</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "      <td>209245.0</td>\n",
       "      <td>12239.0</td>\n",
       "      <td>39.0</td>\n",
       "      <td>27.14</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>bryan</td>\n",
       "      <td>drew</td>\n",
       "      <td>unk</td>\n",
       "      <td>unk</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "      <td>27323.0</td>\n",
       "      <td>12603.5</td>\n",
       "      <td>39.0</td>\n",
       "      <td>13.35</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   surname  forename ticket_str cabin_str   sex embarked  ticket_num  \\\n",
       "0  oconnor   frankie        unk         c  male        s    209245.0   \n",
       "1    bryan      drew        unk       unk  male        s     27323.0   \n",
       "\n",
       "   cabin_num   age   fare  pclass  sibsp  parch  \n",
       "0    12239.0  39.0  27.14       1      2      0  \n",
       "1    12603.5  39.0  13.35       3      0      0  "
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df3.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "cat_features = ['ticket_str', 'cabin_str', 'embarked', 'sex'] + ['surname', 'forename']\n",
    "# text_features = \n",
    "cb_cfg = {\n",
    "    'iterations': 1000,\n",
    "    'task_type': 'CPU',\n",
    "    'cat_features': cat_features,\n",
    "#     'text_features': text_features,\n",
    "    'use_best_model': True,\n",
    "    'early_stopping_rounds':50,\n",
    "    'verbose': True,\n",
    "    'metric_period': 25\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e {color: black;background-color: white;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e pre{padding: 0;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-toggleable {background-color: white;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-estimator:hover {background-color: #d4ebff;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-item {z-index: 1;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-parallel::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-parallel-item:only-child::after {width: 0;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;position: relative;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-label-container {position: relative;z-index: 2;text-align: center;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-e3f787bd-3e72-4b87-ac92-3b1e5377230e\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;pl_simple&#x27;,\n",
       "                 Pipeline(steps=[(&#x27;pl&#x27;,\n",
       "                                  Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                                                   featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                                                   PassNameTransformer()),\n",
       "                                                                                  (&#x27;ticket&#x27;,\n",
       "                                                                                   TicketTransformer()),\n",
       "                                                                                  (&#x27;cabin&#x27;,\n",
       "                                                                                   CabinTransformer()),\n",
       "                                                                                  (&#x27;others&#x27;,\n",
       "                                                                                   ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                                                    &#x27;sex&#x27;,\n",
       "                                                                                                                    &#x27;age&#x27;,\n",
       "                                                                                                                    &#x27;sibsp&#x27;,\n",
       "                                                                                                                    &#x27;parch&#x27;,\n",
       "                                                                                                                    &#x27;fare&#x27;,\n",
       "                                                                                                                    &#x27;embarked&#x27;]))])),\n",
       "                                                  (&#x27;lower_text_values&#x27;,\n",
       "                                                   LowerTran...\n",
       "                                                                   simpleImputer(strategy=&#x27;most_frequent&#x27;),\n",
       "                                                                   &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070&gt;)]))])),\n",
       "                (&#x27;catboost_simple&#x27;,\n",
       "                 CustomCatBoostClassifier(cat_features=[&#x27;ticket_str&#x27;,\n",
       "                                                        &#x27;cabin_str&#x27;, &#x27;embarked&#x27;,\n",
       "                                                        &#x27;sex&#x27;, &#x27;surname&#x27;,\n",
       "                                                        &#x27;forename&#x27;],\n",
       "                                          early_stopping_rounds=50,\n",
       "                                          iterations=1000, metric_period=25,\n",
       "                                          task_type=&#x27;CPU&#x27;, use_best_model=True,\n",
       "                                          verbose=True))])</pre><b>Please rerun this cell to show the HTML repr or trust the notebook.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"eb39e765-90fb-4c3d-a8b4-06a7f16a2f79\" type=\"checkbox\" ><label for=\"eb39e765-90fb-4c3d-a8b4-06a7f16a2f79\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;pl_simple&#x27;,\n",
       "                 Pipeline(steps=[(&#x27;pl&#x27;,\n",
       "                                  Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                                                   featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                                                   PassNameTransformer()),\n",
       "                                                                                  (&#x27;ticket&#x27;,\n",
       "                                                                                   TicketTransformer()),\n",
       "                                                                                  (&#x27;cabin&#x27;,\n",
       "                                                                                   CabinTransformer()),\n",
       "                                                                                  (&#x27;others&#x27;,\n",
       "                                                                                   ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                                                    &#x27;sex&#x27;,\n",
       "                                                                                                                    &#x27;age&#x27;,\n",
       "                                                                                                                    &#x27;sibsp&#x27;,\n",
       "                                                                                                                    &#x27;parch&#x27;,\n",
       "                                                                                                                    &#x27;fare&#x27;,\n",
       "                                                                                                                    &#x27;embarked&#x27;]))])),\n",
       "                                                  (&#x27;lower_text_values&#x27;,\n",
       "                                                   LowerTran...\n",
       "                                                                   simpleImputer(strategy=&#x27;most_frequent&#x27;),\n",
       "                                                                   &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070&gt;)]))])),\n",
       "                (&#x27;catboost_simple&#x27;,\n",
       "                 CustomCatBoostClassifier(cat_features=[&#x27;ticket_str&#x27;,\n",
       "                                                        &#x27;cabin_str&#x27;, &#x27;embarked&#x27;,\n",
       "                                                        &#x27;sex&#x27;, &#x27;surname&#x27;,\n",
       "                                                        &#x27;forename&#x27;],\n",
       "                                          early_stopping_rounds=50,\n",
       "                                          iterations=1000, metric_period=25,\n",
       "                                          task_type=&#x27;CPU&#x27;, use_best_model=True,\n",
       "                                          verbose=True))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"4a0772d1-1f3c-4d78-ae0e-8fc35d1bd4b1\" type=\"checkbox\" ><label for=\"4a0772d1-1f3c-4d78-ae0e-8fc35d1bd4b1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">pl_simple: Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;pl&#x27;,\n",
       "                 Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                                  featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                                  PassNameTransformer()),\n",
       "                                                                 (&#x27;ticket&#x27;,\n",
       "                                                                  TicketTransformer()),\n",
       "                                                                 (&#x27;cabin&#x27;,\n",
       "                                                                  CabinTransformer()),\n",
       "                                                                 (&#x27;others&#x27;,\n",
       "                                                                  ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                                   &#x27;sex&#x27;,\n",
       "                                                                                                   &#x27;age&#x27;,\n",
       "                                                                                                   &#x27;sibsp&#x27;,\n",
       "                                                                                                   &#x27;parch&#x27;,\n",
       "                                                                                                   &#x27;fare&#x27;,\n",
       "                                                                                                   &#x27;embarked&#x27;]))])),\n",
       "                                 (&#x27;lower_text_values&#x27;, LowerTransformer())])),\n",
       "                (&#x27;null_handling...\n",
       "                                                  &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9f10&gt;),\n",
       "                                                 (&#x27;float_features&#x27;,\n",
       "                                                  simpleImputer(strategy=&#x27;median&#x27;),\n",
       "                                                  &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9850&gt;),\n",
       "                                                 (&#x27;count_features&#x27;,\n",
       "                                                  simpleImputer(strategy=&#x27;most_frequent&#x27;),\n",
       "                                                  &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070&gt;)]))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"9b4ee014-30c5-42c1-a8b4-84cbe4b3a098\" type=\"checkbox\" ><label for=\"9b4ee014-30c5-42c1-a8b4-84cbe4b3a098\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">pl: Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                 featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                 PassNameTransformer()),\n",
       "                                                (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                                                (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                                                (&#x27;others&#x27;,\n",
       "                                                 ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                  &#x27;sex&#x27;,\n",
       "                                                                                  &#x27;age&#x27;,\n",
       "                                                                                  &#x27;sibsp&#x27;,\n",
       "                                                                                  &#x27;parch&#x27;,\n",
       "                                                                                  &#x27;fare&#x27;,\n",
       "                                                                                  &#x27;embarked&#x27;]))])),\n",
       "                (&#x27;lower_text_values&#x27;, LowerTransformer())])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"86f352fb-67db-4d91-baae-83b49fd47ea7\" type=\"checkbox\" ><label for=\"86f352fb-67db-4d91-baae-83b49fd47ea7\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">raw_data_processor: featureUnion</label><div class=\"sk-toggleable__content\"><pre>featureUnion(transformer_list=[(&#x27;pass_name&#x27;, PassNameTransformer()),\n",
       "                               (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                               (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                               (&#x27;others&#x27;,\n",
       "                                ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                 &#x27;sex&#x27;, &#x27;age&#x27;,\n",
       "                                                                 &#x27;sibsp&#x27;,\n",
       "                                                                 &#x27;parch&#x27;,\n",
       "                                                                 &#x27;fare&#x27;,\n",
       "                                                                 &#x27;embarked&#x27;]))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>pass_name</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"848782e3-2f50-40ed-b5ea-dd85fdeabba3\" type=\"checkbox\" ><label for=\"848782e3-2f50-40ed-b5ea-dd85fdeabba3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">PassNameTransformer</label><div class=\"sk-toggleable__content\"><pre>PassNameTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>ticket</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"2124ed55-6fdf-4b21-839f-00b3d5cb9751\" type=\"checkbox\" ><label for=\"2124ed55-6fdf-4b21-839f-00b3d5cb9751\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">TicketTransformer</label><div class=\"sk-toggleable__content\"><pre>TicketTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>cabin</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"01fd10a7-75a3-4761-92ae-a614b6e849c5\" type=\"checkbox\" ><label for=\"01fd10a7-75a3-4761-92ae-a614b6e849c5\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CabinTransformer</label><div class=\"sk-toggleable__content\"><pre>CabinTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>others</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"6aebf13c-cb9d-4432-90c6-00e9a4d87a3c\" type=\"checkbox\" ><label for=\"6aebf13c-cb9d-4432-90c6-00e9a4d87a3c\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ColumnSelectTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnSelectTransformer(columns=[&#x27;pclass&#x27;, &#x27;sex&#x27;, &#x27;age&#x27;, &#x27;sibsp&#x27;, &#x27;parch&#x27;,\n",
       "                                 &#x27;fare&#x27;, &#x27;embarked&#x27;])</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"039235ac-806f-4d38-8d5b-496e411c71bc\" type=\"checkbox\" ><label for=\"039235ac-806f-4d38-8d5b-496e411c71bc\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">LowerTransformer</label><div class=\"sk-toggleable__content\"><pre>LowerTransformer()</pre></div></div></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"5c084680-c08f-4a38-93e2-fa652fa0a9f2\" type=\"checkbox\" ><label for=\"5c084680-c08f-4a38-93e2-fa652fa0a9f2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">null_handling: columnTransformer</label><div class=\"sk-toggleable__content\"><pre>columnTransformer(remainder=&#x27;passthrough&#x27;,\n",
       "                  transformers=[(&#x27;text_features&#x27;,\n",
       "                                 simpleImputer(fill_value=&#x27;unk&#x27;,\n",
       "                                               missing_values=None,\n",
       "                                               strategy=&#x27;constant&#x27;),\n",
       "                                 &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9f10&gt;),\n",
       "                                (&#x27;float_features&#x27;,\n",
       "                                 simpleImputer(strategy=&#x27;median&#x27;),\n",
       "                                 &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9850&gt;),\n",
       "                                (&#x27;count_features&#x27;,\n",
       "                                 simpleImputer(strategy=&#x27;most_frequent&#x27;),\n",
       "                                 &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070&gt;)])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"a453bcd1-1908-4918-b153-7a823541ccab\" type=\"checkbox\" ><label for=\"a453bcd1-1908-4918-b153-7a823541ccab\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">text_features</label><div class=\"sk-toggleable__content\"><pre>&lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9f10&gt;</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"a2197e67-5c7a-4674-a2c6-f3cc0fa96261\" type=\"checkbox\" ><label for=\"a2197e67-5c7a-4674-a2c6-f3cc0fa96261\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">simpleImputer</label><div class=\"sk-toggleable__content\"><pre>simpleImputer(fill_value=&#x27;unk&#x27;, missing_values=None, strategy=&#x27;constant&#x27;)</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"cb860a88-36a3-486d-bb7d-6fc5bba54d47\" type=\"checkbox\" ><label for=\"cb860a88-36a3-486d-bb7d-6fc5bba54d47\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">float_features</label><div class=\"sk-toggleable__content\"><pre>&lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9850&gt;</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"bbae15d8-b621-457d-85a5-827290d22634\" type=\"checkbox\" ><label for=\"bbae15d8-b621-457d-85a5-827290d22634\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">simpleImputer</label><div class=\"sk-toggleable__content\"><pre>simpleImputer(strategy=&#x27;median&#x27;)</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"f4f1bd01-d365-4b1a-a939-a75564260342\" type=\"checkbox\" ><label for=\"f4f1bd01-d365-4b1a-a939-a75564260342\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">count_features</label><div class=\"sk-toggleable__content\"><pre>&lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070&gt;</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"18acacb7-7af2-4a0e-9bc9-c3cd3d940dee\" type=\"checkbox\" ><label for=\"18acacb7-7af2-4a0e-9bc9-c3cd3d940dee\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">simpleImputer</label><div class=\"sk-toggleable__content\"><pre>simpleImputer(strategy=&#x27;most_frequent&#x27;)</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"a46a2e89-a8a0-4dfd-9fc8-77c60a3ca25f\" type=\"checkbox\" ><label for=\"a46a2e89-a8a0-4dfd-9fc8-77c60a3ca25f\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">remainder</label><div class=\"sk-toggleable__content\"><pre>[]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"cf890fbb-4f6f-41f2-a82a-9eddc0542ef9\" type=\"checkbox\" ><label for=\"cf890fbb-4f6f-41f2-a82a-9eddc0542ef9\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">passthrough</label><div class=\"sk-toggleable__content\"><pre>passthrough</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"4201f967-ab8f-418b-acc5-444689648db5\" type=\"checkbox\" ><label for=\"4201f967-ab8f-418b-acc5-444689648db5\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CustomCatBoostClassifier</label><div class=\"sk-toggleable__content\"><pre>CustomCatBoostClassifier(cat_features=[&#x27;ticket_str&#x27;, &#x27;cabin_str&#x27;, &#x27;embarked&#x27;,\n",
       "                                       &#x27;sex&#x27;, &#x27;surname&#x27;, &#x27;forename&#x27;],\n",
       "                         early_stopping_rounds=50, iterations=1000,\n",
       "                         metric_period=25, task_type=&#x27;CPU&#x27;, use_best_model=True,\n",
       "                         verbose=True)</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('pl_simple',\n",
       "                 Pipeline(steps=[('pl',\n",
       "                                  Pipeline(steps=[('raw_data_processor',\n",
       "                                                   featureUnion(transformer_list=[('pass_name',\n",
       "                                                                                   PassNameTransformer()),\n",
       "                                                                                  ('ticket',\n",
       "                                                                                   TicketTransformer()),\n",
       "                                                                                  ('cabin',\n",
       "                                                                                   CabinTransformer()),\n",
       "                                                                                  ('others',\n",
       "                                                                                   ColumnSelectTransformer(columns=['pclass',\n",
       "                                                                                                                    'sex',\n",
       "                                                                                                                    'age',\n",
       "                                                                                                                    'sibsp',\n",
       "                                                                                                                    'parch',\n",
       "                                                                                                                    'fare',\n",
       "                                                                                                                    'embarked']))])),\n",
       "                                                  ('lower_text_values',\n",
       "                                                   LowerTran...\n",
       "                                                                   simpleImputer(strategy='most_frequent'),\n",
       "                                                                   <sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070>)]))])),\n",
       "                ('catboost_simple',\n",
       "                 CustomCatBoostClassifier(cat_features=['ticket_str',\n",
       "                                                        'cabin_str', 'embarked',\n",
       "                                                        'sex', 'surname',\n",
       "                                                        'forename'],\n",
       "                                          early_stopping_rounds=50,\n",
       "                                          iterations=1000, metric_period=25,\n",
       "                                          task_type='CPU', use_best_model=True,\n",
       "                                          verbose=True))])"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl4 = Pipeline(steps=[\n",
    "    ('pl_simple', pl_simple),\n",
    "    ('catboost_simple', CustomCatBoostClassifier(**cb_cfg))\n",
    "])\n",
    "pl4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e932a8a3e03147149319a04025f5e799",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b3e3d3931c5b45038e75180cc1a0d3eb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f97ab346e4bb4f1ca96e4dac8e4f82af",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d91919b8f91e424ba0d898782c7fc8e6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[['surname', 'forename', 'ticket_str', 'cabin_str', 'sex', 'embarked'], ['ticket_num', 'cabin_num', 'age', 'fare'], ['pclass', 'sibsp', 'parch']]\n",
      "[(100000, 6), (100000, 4), (100000, 3)]\n",
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "64dd0e70906447a49d344f09f20d4217",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "======== CustomCatBoostClassifier - START ========\n",
      "Learning rate set to 0.096296\n",
      "0:\tlearn: 0.6509106\ttest: 0.6500870\tbest: 0.6500870 (0)\ttotal: 114ms\tremaining: 1m 53s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25:\tlearn: 0.4771816\ttest: 0.4732935\tbest: 0.4732935 (25)\ttotal: 998ms\tremaining: 37.4s\n",
      "50:\tlearn: 0.4697525\ttest: 0.4668533\tbest: 0.4668533 (50)\ttotal: 1.81s\tremaining: 33.7s\n",
      "75:\tlearn: 0.4667908\ttest: 0.4652342\tbest: 0.4652342 (75)\ttotal: 2.62s\tremaining: 31.9s\n",
      "100:\tlearn: 0.4649027\ttest: 0.4643844\tbest: 0.4643844 (100)\ttotal: 3.47s\tremaining: 30.9s\n",
      "125:\tlearn: 0.4630746\ttest: 0.4636356\tbest: 0.4636172 (124)\ttotal: 4.19s\tremaining: 29.1s\n",
      "150:\tlearn: 0.4609647\ttest: 0.4626024\tbest: 0.4626024 (150)\ttotal: 4.97s\tremaining: 28s\n",
      "175:\tlearn: 0.4595652\ttest: 0.4623817\tbest: 0.4623743 (174)\ttotal: 5.72s\tremaining: 26.8s\n",
      "200:\tlearn: 0.4583941\ttest: 0.4622576\tbest: 0.4622420 (199)\ttotal: 6.51s\tremaining: 25.9s\n",
      "225:\tlearn: 0.4571629\ttest: 0.4620526\tbest: 0.4620526 (225)\ttotal: 7.28s\tremaining: 24.9s\n",
      "250:\tlearn: 0.4560363\ttest: 0.4619599\tbest: 0.4619213 (241)\ttotal: 8.05s\tremaining: 24s\n",
      "275:\tlearn: 0.4547426\ttest: 0.4618220\tbest: 0.4617892 (274)\ttotal: 8.82s\tremaining: 23.1s\n",
      "300:\tlearn: 0.4536035\ttest: 0.4615407\tbest: 0.4615407 (300)\ttotal: 9.69s\tremaining: 22.5s\n",
      "325:\tlearn: 0.4525477\ttest: 0.4615069\tbest: 0.4614089 (320)\ttotal: 10.5s\tremaining: 21.7s\n",
      "350:\tlearn: 0.4514910\ttest: 0.4617423\tbest: 0.4614089 (320)\ttotal: 11.4s\tremaining: 21s\n",
      "Stopped by overfitting detector  (50 iterations wait)\n",
      "\n",
      "bestTest = 0.4614088654\n",
      "bestIteration = 320\n",
      "\n",
      "Shrink model to first 321 iterations.\n",
      "Evaluating model\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.81      0.83      0.82      5762\n",
      "           1       0.76      0.74      0.75      4238\n",
      "\n",
      "    accuracy                           0.79     10000\n",
      "   macro avg       0.78      0.78      0.78     10000\n",
      "weighted avg       0.79      0.79      0.79     10000\n",
      "\n",
      "0: 0.8594034175184636\n",
      "1: 0.8594034175184636\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "======== CustomCatBoostClassifier - FINISH =======> Take: 12.573838(s)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style>#sk-124a1bc9-08dc-4871-a8a0-59046282597c {color: black;background-color: white;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c pre{padding: 0;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-toggleable {background-color: white;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-estimator:hover {background-color: #d4ebff;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-item {z-index: 1;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-parallel::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-parallel-item:only-child::after {width: 0;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;position: relative;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-label-container {position: relative;z-index: 2;text-align: center;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-124a1bc9-08dc-4871-a8a0-59046282597c div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-124a1bc9-08dc-4871-a8a0-59046282597c\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;pl_simple&#x27;,\n",
       "                 Pipeline(steps=[(&#x27;pl&#x27;,\n",
       "                                  Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                                                   featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                                                   PassNameTransformer()),\n",
       "                                                                                  (&#x27;ticket&#x27;,\n",
       "                                                                                   TicketTransformer()),\n",
       "                                                                                  (&#x27;cabin&#x27;,\n",
       "                                                                                   CabinTransformer()),\n",
       "                                                                                  (&#x27;others&#x27;,\n",
       "                                                                                   ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                                                    &#x27;sex&#x27;,\n",
       "                                                                                                                    &#x27;age&#x27;,\n",
       "                                                                                                                    &#x27;sibsp&#x27;,\n",
       "                                                                                                                    &#x27;parch&#x27;,\n",
       "                                                                                                                    &#x27;fare&#x27;,\n",
       "                                                                                                                    &#x27;embarked&#x27;]))])),\n",
       "                                                  (&#x27;lower_text_values&#x27;,\n",
       "                                                   LowerTran...\n",
       "                                                                   simpleImputer(strategy=&#x27;most_frequent&#x27;),\n",
       "                                                                   &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070&gt;)]))])),\n",
       "                (&#x27;catboost_simple&#x27;,\n",
       "                 CustomCatBoostClassifier(cat_features=[&#x27;ticket_str&#x27;,\n",
       "                                                        &#x27;cabin_str&#x27;, &#x27;embarked&#x27;,\n",
       "                                                        &#x27;sex&#x27;, &#x27;surname&#x27;,\n",
       "                                                        &#x27;forename&#x27;],\n",
       "                                          early_stopping_rounds=50,\n",
       "                                          iterations=1000, metric_period=25,\n",
       "                                          task_type=&#x27;CPU&#x27;, use_best_model=True,\n",
       "                                          verbose=True))])</pre><b>Please rerun this cell to show the HTML repr or trust the notebook.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"6d13ea00-b308-491d-807b-574e90e4af4e\" type=\"checkbox\" ><label for=\"6d13ea00-b308-491d-807b-574e90e4af4e\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;pl_simple&#x27;,\n",
       "                 Pipeline(steps=[(&#x27;pl&#x27;,\n",
       "                                  Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                                                   featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                                                   PassNameTransformer()),\n",
       "                                                                                  (&#x27;ticket&#x27;,\n",
       "                                                                                   TicketTransformer()),\n",
       "                                                                                  (&#x27;cabin&#x27;,\n",
       "                                                                                   CabinTransformer()),\n",
       "                                                                                  (&#x27;others&#x27;,\n",
       "                                                                                   ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                                                    &#x27;sex&#x27;,\n",
       "                                                                                                                    &#x27;age&#x27;,\n",
       "                                                                                                                    &#x27;sibsp&#x27;,\n",
       "                                                                                                                    &#x27;parch&#x27;,\n",
       "                                                                                                                    &#x27;fare&#x27;,\n",
       "                                                                                                                    &#x27;embarked&#x27;]))])),\n",
       "                                                  (&#x27;lower_text_values&#x27;,\n",
       "                                                   LowerTran...\n",
       "                                                                   simpleImputer(strategy=&#x27;most_frequent&#x27;),\n",
       "                                                                   &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070&gt;)]))])),\n",
       "                (&#x27;catboost_simple&#x27;,\n",
       "                 CustomCatBoostClassifier(cat_features=[&#x27;ticket_str&#x27;,\n",
       "                                                        &#x27;cabin_str&#x27;, &#x27;embarked&#x27;,\n",
       "                                                        &#x27;sex&#x27;, &#x27;surname&#x27;,\n",
       "                                                        &#x27;forename&#x27;],\n",
       "                                          early_stopping_rounds=50,\n",
       "                                          iterations=1000, metric_period=25,\n",
       "                                          task_type=&#x27;CPU&#x27;, use_best_model=True,\n",
       "                                          verbose=True))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"18d010aa-48ae-4b48-b277-4b0a719bd0dc\" type=\"checkbox\" ><label for=\"18d010aa-48ae-4b48-b277-4b0a719bd0dc\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">pl_simple: Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;pl&#x27;,\n",
       "                 Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                                  featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                                  PassNameTransformer()),\n",
       "                                                                 (&#x27;ticket&#x27;,\n",
       "                                                                  TicketTransformer()),\n",
       "                                                                 (&#x27;cabin&#x27;,\n",
       "                                                                  CabinTransformer()),\n",
       "                                                                 (&#x27;others&#x27;,\n",
       "                                                                  ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                                   &#x27;sex&#x27;,\n",
       "                                                                                                   &#x27;age&#x27;,\n",
       "                                                                                                   &#x27;sibsp&#x27;,\n",
       "                                                                                                   &#x27;parch&#x27;,\n",
       "                                                                                                   &#x27;fare&#x27;,\n",
       "                                                                                                   &#x27;embarked&#x27;]))])),\n",
       "                                 (&#x27;lower_text_values&#x27;, LowerTransformer())])),\n",
       "                (&#x27;null_handling...\n",
       "                                                  &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9f10&gt;),\n",
       "                                                 (&#x27;float_features&#x27;,\n",
       "                                                  simpleImputer(strategy=&#x27;median&#x27;),\n",
       "                                                  &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9850&gt;),\n",
       "                                                 (&#x27;count_features&#x27;,\n",
       "                                                  simpleImputer(strategy=&#x27;most_frequent&#x27;),\n",
       "                                                  &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070&gt;)]))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"80c730e5-bf39-4dea-b483-52a6c8d0dbc0\" type=\"checkbox\" ><label for=\"80c730e5-bf39-4dea-b483-52a6c8d0dbc0\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">pl: Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                 featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                 PassNameTransformer()),\n",
       "                                                (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                                                (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                                                (&#x27;others&#x27;,\n",
       "                                                 ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                  &#x27;sex&#x27;,\n",
       "                                                                                  &#x27;age&#x27;,\n",
       "                                                                                  &#x27;sibsp&#x27;,\n",
       "                                                                                  &#x27;parch&#x27;,\n",
       "                                                                                  &#x27;fare&#x27;,\n",
       "                                                                                  &#x27;embarked&#x27;]))])),\n",
       "                (&#x27;lower_text_values&#x27;, LowerTransformer())])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"baca2025-abd8-41f5-9d75-f42afed8274b\" type=\"checkbox\" ><label for=\"baca2025-abd8-41f5-9d75-f42afed8274b\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">raw_data_processor: featureUnion</label><div class=\"sk-toggleable__content\"><pre>featureUnion(transformer_list=[(&#x27;pass_name&#x27;, PassNameTransformer()),\n",
       "                               (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                               (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                               (&#x27;others&#x27;,\n",
       "                                ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                 &#x27;sex&#x27;, &#x27;age&#x27;,\n",
       "                                                                 &#x27;sibsp&#x27;,\n",
       "                                                                 &#x27;parch&#x27;,\n",
       "                                                                 &#x27;fare&#x27;,\n",
       "                                                                 &#x27;embarked&#x27;]))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>pass_name</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"b99faa9a-3a4e-40d8-b06e-4ba1957ba07e\" type=\"checkbox\" ><label for=\"b99faa9a-3a4e-40d8-b06e-4ba1957ba07e\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">PassNameTransformer</label><div class=\"sk-toggleable__content\"><pre>PassNameTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>ticket</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"1f2dfcf0-297d-4591-ac90-c729e12f2d28\" type=\"checkbox\" ><label for=\"1f2dfcf0-297d-4591-ac90-c729e12f2d28\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">TicketTransformer</label><div class=\"sk-toggleable__content\"><pre>TicketTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>cabin</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"193b4560-4d01-48b1-888a-e2b1efcd1827\" type=\"checkbox\" ><label for=\"193b4560-4d01-48b1-888a-e2b1efcd1827\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CabinTransformer</label><div class=\"sk-toggleable__content\"><pre>CabinTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>others</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"a290be37-e8a0-4dba-9e99-c7f189945781\" type=\"checkbox\" ><label for=\"a290be37-e8a0-4dba-9e99-c7f189945781\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ColumnSelectTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnSelectTransformer(columns=[&#x27;pclass&#x27;, &#x27;sex&#x27;, &#x27;age&#x27;, &#x27;sibsp&#x27;, &#x27;parch&#x27;,\n",
       "                                 &#x27;fare&#x27;, &#x27;embarked&#x27;])</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"d85eb47d-5fd1-4cf8-95e7-5072d4cd049f\" type=\"checkbox\" ><label for=\"d85eb47d-5fd1-4cf8-95e7-5072d4cd049f\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">LowerTransformer</label><div class=\"sk-toggleable__content\"><pre>LowerTransformer()</pre></div></div></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"65ea923e-c1e4-44e1-b769-c2b881c3b470\" type=\"checkbox\" ><label for=\"65ea923e-c1e4-44e1-b769-c2b881c3b470\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">null_handling: columnTransformer</label><div class=\"sk-toggleable__content\"><pre>columnTransformer(remainder=&#x27;passthrough&#x27;,\n",
       "                  transformers=[(&#x27;text_features&#x27;,\n",
       "                                 simpleImputer(fill_value=&#x27;unk&#x27;,\n",
       "                                               missing_values=None,\n",
       "                                               strategy=&#x27;constant&#x27;),\n",
       "                                 &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9f10&gt;),\n",
       "                                (&#x27;float_features&#x27;,\n",
       "                                 simpleImputer(strategy=&#x27;median&#x27;),\n",
       "                                 &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9850&gt;),\n",
       "                                (&#x27;count_features&#x27;,\n",
       "                                 simpleImputer(strategy=&#x27;most_frequent&#x27;),\n",
       "                                 &lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070&gt;)])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"9558e21a-9c86-47d6-9877-13b97a237c0f\" type=\"checkbox\" ><label for=\"9558e21a-9c86-47d6-9877-13b97a237c0f\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">text_features</label><div class=\"sk-toggleable__content\"><pre>&lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9f10&gt;</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"6a4be3da-d944-4a40-8435-8f53c2491d42\" type=\"checkbox\" ><label for=\"6a4be3da-d944-4a40-8435-8f53c2491d42\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">simpleImputer</label><div class=\"sk-toggleable__content\"><pre>simpleImputer(fill_value=&#x27;unk&#x27;, missing_values=None, strategy=&#x27;constant&#x27;)</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"4fba8a26-d709-4a36-82a7-b8d42f1a0c3f\" type=\"checkbox\" ><label for=\"4fba8a26-d709-4a36-82a7-b8d42f1a0c3f\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">float_features</label><div class=\"sk-toggleable__content\"><pre>&lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9850&gt;</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"c790e252-7dea-4197-99b9-4994ff083d0d\" type=\"checkbox\" ><label for=\"c790e252-7dea-4197-99b9-4994ff083d0d\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">simpleImputer</label><div class=\"sk-toggleable__content\"><pre>simpleImputer(strategy=&#x27;median&#x27;)</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"ca6fc259-0b2f-4d8f-b4ad-7782b766ad58\" type=\"checkbox\" ><label for=\"ca6fc259-0b2f-4d8f-b4ad-7782b766ad58\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">count_features</label><div class=\"sk-toggleable__content\"><pre>&lt;sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070&gt;</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"9a38f09c-60d6-423b-ba31-e898e6b74cd0\" type=\"checkbox\" ><label for=\"9a38f09c-60d6-423b-ba31-e898e6b74cd0\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">simpleImputer</label><div class=\"sk-toggleable__content\"><pre>simpleImputer(strategy=&#x27;most_frequent&#x27;)</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"1364e323-b332-4076-ab64-214f5f3ae71b\" type=\"checkbox\" ><label for=\"1364e323-b332-4076-ab64-214f5f3ae71b\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">remainder</label><div class=\"sk-toggleable__content\"><pre>[]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"5c5b8de2-58a6-4521-abb5-78f444181d2e\" type=\"checkbox\" ><label for=\"5c5b8de2-58a6-4521-abb5-78f444181d2e\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">passthrough</label><div class=\"sk-toggleable__content\"><pre>passthrough</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"5dcaa8e6-5c6d-462c-a79c-b7f28950d003\" type=\"checkbox\" ><label for=\"5dcaa8e6-5c6d-462c-a79c-b7f28950d003\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CustomCatBoostClassifier</label><div class=\"sk-toggleable__content\"><pre>CustomCatBoostClassifier(cat_features=[&#x27;ticket_str&#x27;, &#x27;cabin_str&#x27;, &#x27;embarked&#x27;,\n",
       "                                       &#x27;sex&#x27;, &#x27;surname&#x27;, &#x27;forename&#x27;],\n",
       "                         early_stopping_rounds=50, iterations=1000,\n",
       "                         metric_period=25, task_type=&#x27;CPU&#x27;, use_best_model=True,\n",
       "                         verbose=True)</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('pl_simple',\n",
       "                 Pipeline(steps=[('pl',\n",
       "                                  Pipeline(steps=[('raw_data_processor',\n",
       "                                                   featureUnion(transformer_list=[('pass_name',\n",
       "                                                                                   PassNameTransformer()),\n",
       "                                                                                  ('ticket',\n",
       "                                                                                   TicketTransformer()),\n",
       "                                                                                  ('cabin',\n",
       "                                                                                   CabinTransformer()),\n",
       "                                                                                  ('others',\n",
       "                                                                                   ColumnSelectTransformer(columns=['pclass',\n",
       "                                                                                                                    'sex',\n",
       "                                                                                                                    'age',\n",
       "                                                                                                                    'sibsp',\n",
       "                                                                                                                    'parch',\n",
       "                                                                                                                    'fare',\n",
       "                                                                                                                    'embarked']))])),\n",
       "                                                  ('lower_text_values',\n",
       "                                                   LowerTran...\n",
       "                                                                   simpleImputer(strategy='most_frequent'),\n",
       "                                                                   <sklearn.compose._column_transformer.make_column_selector object at 0x7f84d8ce9070>)]))])),\n",
       "                ('catboost_simple',\n",
       "                 CustomCatBoostClassifier(cat_features=['ticket_str',\n",
       "                                                        'cabin_str', 'embarked',\n",
       "                                                        'sex', 'surname',\n",
       "                                                        'forename'],\n",
       "                                          early_stopping_rounds=50,\n",
       "                                          iterations=1000, metric_period=25,\n",
       "                                          task_type='CPU', use_best_model=True,\n",
       "                                          verbose=True))])"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl4.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c41b62e25b6940b3a8d52fdfb3bae340",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6fddab7677e5488eb85f0328ec6c0f26",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a100678b5fee4e2f90c3471d3edd26bf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b1956854f59742f38b0787f80993c63f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[['surname', 'forename', 'ticket_str', 'cabin_str', 'sex', 'embarked'], ['ticket_num', 'cabin_num', 'age', 'fare'], ['pclass', 'sibsp', 'parch']]\n",
      "[(100000, 6), (100000, 4), (100000, 3)]\n",
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "79290028b241439eafa5ac769f1cdb15",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "b = pl4.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>surname</th>\n",
       "      <th>forename</th>\n",
       "      <th>ticket_str</th>\n",
       "      <th>cabin_str</th>\n",
       "      <th>sex</th>\n",
       "      <th>embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>oconnor</td>\n",
       "      <td>frankie</td>\n",
       "      <td>None</td>\n",
       "      <td>c</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>bryan</td>\n",
       "      <td>drew</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>owens</td>\n",
       "      <td>kenneth</td>\n",
       "      <td>ca</td>\n",
       "      <td>None</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>kramer</td>\n",
       "      <td>james</td>\n",
       "      <td>a.</td>\n",
       "      <td>None</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>bond</td>\n",
       "      <td>michael</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>male</td>\n",
       "      <td>s</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99995</th>\n",
       "      <td>bell</td>\n",
       "      <td>adele</td>\n",
       "      <td>pc</td>\n",
       "      <td>d</td>\n",
       "      <td>female</td>\n",
       "      <td>c</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99996</th>\n",
       "      <td>brown</td>\n",
       "      <td>herman</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99997</th>\n",
       "      <td>childress</td>\n",
       "      <td>charles</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99998</th>\n",
       "      <td>caughlin</td>\n",
       "      <td>thomas</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99999</th>\n",
       "      <td>enciso</td>\n",
       "      <td>tyler</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>male</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100000 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         surname  forename ticket_str cabin_str     sex embarked\n",
       "0        oconnor   frankie       None         c    male        s\n",
       "1          bryan      drew       None      None    male        s\n",
       "2          owens   kenneth         ca      None    male        s\n",
       "3         kramer     james         a.      None    male        s\n",
       "4           bond   michael       None      None    male        s\n",
       "...          ...       ...        ...       ...     ...      ...\n",
       "99995       bell     adele         pc         d  female        c\n",
       "99996      brown    herman       None      None    male        s\n",
       "99997  childress   charles       None      None    male        s\n",
       "99998   caughlin    thomas       None      None    male        s\n",
       "99999     enciso     tyler       None      None    male        s\n",
       "\n",
       "[100000 rows x 6 columns]"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2[['surname', 'forename', 'ticket_str', 'cabin_str', 'sex', 'embarked']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import LabelEncoder\n",
    "class CustomLabelEncoder(PipelineLogger,TransformerMixin, BaseEstimator):\n",
    "    def __init__(self, columns=None):\n",
    "        super(CustomLabelEncoder).__init__()\n",
    "        self.columns = columns\n",
    "        self.label_encoders = {}\n",
    "        \n",
    "    def fit(self, X, y=None):\n",
    "        X_ = X[self.columns]\n",
    "        X_ = pd.concat([X_, pd.DataFrame([['unk'] * X_.shape[1]], columns=X_.columns)], ignore_index=True)\n",
    "        for c in self.columns:\n",
    "            self.label_encoders[c] = LabelEncoder().fit(X_[c])\n",
    "        print(self.label_encoders)\n",
    "        return self\n",
    "    \n",
    "    def transform(self, X, y=None):\n",
    "        X_ = pd.DataFrame()\n",
    "        self.dummy_dicts = {}\n",
    "        for c in self.columns:\n",
    "            dd = {}\n",
    "            for cl in self.label_encoders[c].classes_:\n",
    "                dd[cl] = True\n",
    "            sr = X[c].map(lambda s: 'unk' if dd.get(s) is None else s)\n",
    "            X_[c] = self.label_encoders[c].transform(sr)\n",
    "        return X_\n",
    "            \n",
    "    def inverse_transform(self, X_encode, y=None):\n",
    "        X_decode = pd.DataFrame()\n",
    "        for c in self.columns:\n",
    "            X_decode[c] = self.label_encoders[c].inverse_transform(X_encode[c])\n",
    "        return X_decode"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "mul_le = CustomLabelEncoder(\n",
    "        columns=['surname', 'forename', 'ticket_str', 'cabin_str', 'sex', 'embarked']\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "label_encoding = columnTransformer(transformers=[\n",
    "    ('category_encoder', mul_le, ['surname', 'forename', 'ticket_str', 'cabin_str', 'sex', 'embarked'])\n",
    "], remainder='passthrough')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'surname': LabelEncoder(), 'forename': LabelEncoder(), 'ticket_str': LabelEncoder(), 'cabin_str': LabelEncoder(), 'sex': LabelEncoder(), 'embarked': LabelEncoder()}\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>surname</th>\n",
       "      <th>forename</th>\n",
       "      <th>ticket_str</th>\n",
       "      <th>cabin_str</th>\n",
       "      <th>sex</th>\n",
       "      <th>embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>14195</td>\n",
       "      <td>1311</td>\n",
       "      <td>49</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2495</td>\n",
       "      <td>1015</td>\n",
       "      <td>49</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>14478</td>\n",
       "      <td>2010</td>\n",
       "      <td>14</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10435</td>\n",
       "      <td>1662</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1902</td>\n",
       "      <td>2678</td>\n",
       "      <td>49</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>2</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99995</th>\n",
       "      <td>1301</td>\n",
       "      <td>18</td>\n",
       "      <td>21</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99996</th>\n",
       "      <td>2436</td>\n",
       "      <td>1507</td>\n",
       "      <td>49</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99997</th>\n",
       "      <td>3463</td>\n",
       "      <td>610</td>\n",
       "      <td>49</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99998</th>\n",
       "      <td>3226</td>\n",
       "      <td>3636</td>\n",
       "      <td>49</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99999</th>\n",
       "      <td>5762</td>\n",
       "      <td>3730</td>\n",
       "      <td>49</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100000 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       surname  forename  ticket_str  cabin_str  sex  embarked\n",
       "0        14195      1311          49          2    1         2\n",
       "1         2495      1015          49          9    1         2\n",
       "2        14478      2010          14          9    1         2\n",
       "3        10435      1662           0          9    1         2\n",
       "4         1902      2678          49          9    1         2\n",
       "...        ...       ...         ...        ...  ...       ...\n",
       "99995     1301        18          21          3    0         0\n",
       "99996     2436      1507          49          9    1         2\n",
       "99997     3463       610          49          9    1         2\n",
       "99998     3226      3636          49          9    1         2\n",
       "99999     5762      3730          49          9    1         2\n",
       "\n",
       "[100000 rows x 6 columns]"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mul_le.fit_transform(df2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>surname</th>\n",
       "      <th>forename</th>\n",
       "      <th>ticket_str</th>\n",
       "      <th>ticket_num</th>\n",
       "      <th>cabin_str</th>\n",
       "      <th>cabin_num</th>\n",
       "      <th>pclass</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>sibsp</th>\n",
       "      <th>parch</th>\n",
       "      <th>fare</th>\n",
       "      <th>embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>oconnor</td>\n",
       "      <td>frankie</td>\n",
       "      <td>None</td>\n",
       "      <td>209245.0</td>\n",
       "      <td>c</td>\n",
       "      <td>12239.0</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>27.14</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>bryan</td>\n",
       "      <td>drew</td>\n",
       "      <td>None</td>\n",
       "      <td>27323.0</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13.35</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>owens</td>\n",
       "      <td>kenneth</td>\n",
       "      <td>ca</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>0.33</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>71.29</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>kramer</td>\n",
       "      <td>james</td>\n",
       "      <td>a.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>19.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13.04</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>bond</td>\n",
       "      <td>michael</td>\n",
       "      <td>None</td>\n",
       "      <td>427635.0</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>25.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.76</td>\n",
       "      <td>s</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99995</th>\n",
       "      <td>bell</td>\n",
       "      <td>adele</td>\n",
       "      <td>pc</td>\n",
       "      <td>NaN</td>\n",
       "      <td>d</td>\n",
       "      <td>17243.0</td>\n",
       "      <td>2</td>\n",
       "      <td>female</td>\n",
       "      <td>62.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>14.86</td>\n",
       "      <td>c</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99996</th>\n",
       "      <td>brown</td>\n",
       "      <td>herman</td>\n",
       "      <td>None</td>\n",
       "      <td>13273.0</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>male</td>\n",
       "      <td>66.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>11.15</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99997</th>\n",
       "      <td>childress</td>\n",
       "      <td>charles</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>37.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>9.95</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99998</th>\n",
       "      <td>caughlin</td>\n",
       "      <td>thomas</td>\n",
       "      <td>None</td>\n",
       "      <td>458654.0</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>51.00</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>30.92</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99999</th>\n",
       "      <td>enciso</td>\n",
       "      <td>tyler</td>\n",
       "      <td>None</td>\n",
       "      <td>458074.0</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>55.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13.96</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100000 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         surname  forename ticket_str  ticket_num cabin_str  cabin_num  \\\n",
       "0        oconnor   frankie       None    209245.0         c    12239.0   \n",
       "1          bryan      drew       None     27323.0      None        NaN   \n",
       "2          owens   kenneth         ca         NaN      None        NaN   \n",
       "3         kramer     james         a.         NaN      None        NaN   \n",
       "4           bond   michael       None    427635.0      None        NaN   \n",
       "...          ...       ...        ...         ...       ...        ...   \n",
       "99995       bell     adele         pc         NaN         d    17243.0   \n",
       "99996      brown    herman       None     13273.0      None        NaN   \n",
       "99997  childress   charles       None         NaN      None        NaN   \n",
       "99998   caughlin    thomas       None    458654.0      None        NaN   \n",
       "99999     enciso     tyler       None    458074.0      None        NaN   \n",
       "\n",
       "       pclass     sex    age  sibsp  parch   fare embarked  \n",
       "0           1    male    NaN      2      0  27.14        s  \n",
       "1           3    male    NaN      0      0  13.35        s  \n",
       "2           3    male   0.33      1      2  71.29        s  \n",
       "3           3    male  19.00      0      0  13.04        s  \n",
       "4           3    male  25.00      0      0   7.76        s  \n",
       "...       ...     ...    ...    ...    ...    ...      ...  \n",
       "99995       2  female  62.00      0      0  14.86        c  \n",
       "99996       2    male  66.00      0      0  11.15        s  \n",
       "99997       3    male  37.00      0      0   9.95        s  \n",
       "99998       3    male  51.00      0      1  30.92        s  \n",
       "99999       3    male  55.00      0      0  13.96        s  \n",
       "\n",
       "[100000 rows x 13 columns]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "class CustomLGBMClassifier(ExperimentBase, LGBMClassifier):\n",
    "    pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "pl_lgbm = Pipeline(steps=[\n",
    "    ('prepro', pl),\n",
    "    ('label_encoding', label_encoding),\n",
    "    ('lgbm', CustomLGBMClassifier())\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ecac54f5e6c44262a847606e2e13e982",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'surname': LabelEncoder(), 'forename': LabelEncoder(), 'ticket_str': LabelEncoder(), 'cabin_str': LabelEncoder(), 'sex': LabelEncoder(), 'embarked': LabelEncoder()}\n",
      "[['surname', 'forename', 'ticket_str', 'cabin_str', 'sex', 'embarked'], ['ticket_num', 'cabin_num', 'pclass', 'age', 'sibsp', 'parch', 'fare']]\n",
      "[(100000, 6), (100000, 7)]\n",
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9b5f43c362d540ffa78319a990cb123a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style>#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 {color: black;background-color: white;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 pre{padding: 0;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-toggleable {background-color: white;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-estimator:hover {background-color: #d4ebff;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-item {z-index: 1;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-parallel::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-parallel-item:only-child::after {width: 0;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;position: relative;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-label-container {position: relative;z-index: 2;text-align: center;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-95f9d9ce-fc27-4a78-979b-2ee527525e27 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-95f9d9ce-fc27-4a78-979b-2ee527525e27\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;prepro&#x27;,\n",
       "                 Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                                  featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                                  PassNameTransformer()),\n",
       "                                                                 (&#x27;ticket&#x27;,\n",
       "                                                                  TicketTransformer()),\n",
       "                                                                 (&#x27;cabin&#x27;,\n",
       "                                                                  CabinTransformer()),\n",
       "                                                                 (&#x27;others&#x27;,\n",
       "                                                                  ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                                   &#x27;sex&#x27;,\n",
       "                                                                                                   &#x27;age&#x27;,\n",
       "                                                                                                   &#x27;sibsp&#x27;,\n",
       "                                                                                                   &#x27;parch&#x27;,\n",
       "                                                                                                   &#x27;fare&#x27;,\n",
       "                                                                                                   &#x27;embarked&#x27;]))])),\n",
       "                                 (&#x27;lower_text_values&#x27;, LowerTransformer())])),\n",
       "                (&#x27;label_encoding&#x27;,\n",
       "                 columnTransformer(remainder=&#x27;passthrough&#x27;,\n",
       "                                   transformers=[(&#x27;category_encoder&#x27;,\n",
       "                                                  CustomLabelEncoder(columns=[&#x27;surname&#x27;,\n",
       "                                                                              &#x27;forename&#x27;,\n",
       "                                                                              &#x27;ticket_str&#x27;,\n",
       "                                                                              &#x27;cabin_str&#x27;,\n",
       "                                                                              &#x27;sex&#x27;,\n",
       "                                                                              &#x27;embarked&#x27;]),\n",
       "                                                  [&#x27;surname&#x27;, &#x27;forename&#x27;,\n",
       "                                                   &#x27;ticket_str&#x27;, &#x27;cabin_str&#x27;,\n",
       "                                                   &#x27;sex&#x27;, &#x27;embarked&#x27;])])),\n",
       "                (&#x27;lgbm&#x27;, CustomLGBMClassifier())])</pre><b>Please rerun this cell to show the HTML repr or trust the notebook.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"dbc7622d-af0f-4389-8d8e-cb3650f803e1\" type=\"checkbox\" ><label for=\"dbc7622d-af0f-4389-8d8e-cb3650f803e1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;prepro&#x27;,\n",
       "                 Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                                  featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                                  PassNameTransformer()),\n",
       "                                                                 (&#x27;ticket&#x27;,\n",
       "                                                                  TicketTransformer()),\n",
       "                                                                 (&#x27;cabin&#x27;,\n",
       "                                                                  CabinTransformer()),\n",
       "                                                                 (&#x27;others&#x27;,\n",
       "                                                                  ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                                   &#x27;sex&#x27;,\n",
       "                                                                                                   &#x27;age&#x27;,\n",
       "                                                                                                   &#x27;sibsp&#x27;,\n",
       "                                                                                                   &#x27;parch&#x27;,\n",
       "                                                                                                   &#x27;fare&#x27;,\n",
       "                                                                                                   &#x27;embarked&#x27;]))])),\n",
       "                                 (&#x27;lower_text_values&#x27;, LowerTransformer())])),\n",
       "                (&#x27;label_encoding&#x27;,\n",
       "                 columnTransformer(remainder=&#x27;passthrough&#x27;,\n",
       "                                   transformers=[(&#x27;category_encoder&#x27;,\n",
       "                                                  CustomLabelEncoder(columns=[&#x27;surname&#x27;,\n",
       "                                                                              &#x27;forename&#x27;,\n",
       "                                                                              &#x27;ticket_str&#x27;,\n",
       "                                                                              &#x27;cabin_str&#x27;,\n",
       "                                                                              &#x27;sex&#x27;,\n",
       "                                                                              &#x27;embarked&#x27;]),\n",
       "                                                  [&#x27;surname&#x27;, &#x27;forename&#x27;,\n",
       "                                                   &#x27;ticket_str&#x27;, &#x27;cabin_str&#x27;,\n",
       "                                                   &#x27;sex&#x27;, &#x27;embarked&#x27;])])),\n",
       "                (&#x27;lgbm&#x27;, CustomLGBMClassifier())])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"4375432f-af9c-4b05-8827-62abc0ccc92b\" type=\"checkbox\" ><label for=\"4375432f-af9c-4b05-8827-62abc0ccc92b\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">prepro: Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                 featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                 PassNameTransformer()),\n",
       "                                                (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                                                (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                                                (&#x27;others&#x27;,\n",
       "                                                 ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                  &#x27;sex&#x27;,\n",
       "                                                                                  &#x27;age&#x27;,\n",
       "                                                                                  &#x27;sibsp&#x27;,\n",
       "                                                                                  &#x27;parch&#x27;,\n",
       "                                                                                  &#x27;fare&#x27;,\n",
       "                                                                                  &#x27;embarked&#x27;]))])),\n",
       "                (&#x27;lower_text_values&#x27;, LowerTransformer())])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"e47dc707-1f76-490a-889d-0b8ca9b32af6\" type=\"checkbox\" ><label for=\"e47dc707-1f76-490a-889d-0b8ca9b32af6\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">raw_data_processor: featureUnion</label><div class=\"sk-toggleable__content\"><pre>featureUnion(transformer_list=[(&#x27;pass_name&#x27;, PassNameTransformer()),\n",
       "                               (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                               (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                               (&#x27;others&#x27;,\n",
       "                                ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                 &#x27;sex&#x27;, &#x27;age&#x27;,\n",
       "                                                                 &#x27;sibsp&#x27;,\n",
       "                                                                 &#x27;parch&#x27;,\n",
       "                                                                 &#x27;fare&#x27;,\n",
       "                                                                 &#x27;embarked&#x27;]))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>pass_name</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"fed2c15d-2e80-4acc-8374-a66d13254e4b\" type=\"checkbox\" ><label for=\"fed2c15d-2e80-4acc-8374-a66d13254e4b\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">PassNameTransformer</label><div class=\"sk-toggleable__content\"><pre>PassNameTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>ticket</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"183c657a-9078-4807-8e26-8b80a48c5d44\" type=\"checkbox\" ><label for=\"183c657a-9078-4807-8e26-8b80a48c5d44\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">TicketTransformer</label><div class=\"sk-toggleable__content\"><pre>TicketTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>cabin</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"1e5092d5-7d76-42b3-9825-64eb0ddabc7e\" type=\"checkbox\" ><label for=\"1e5092d5-7d76-42b3-9825-64eb0ddabc7e\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CabinTransformer</label><div class=\"sk-toggleable__content\"><pre>CabinTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>others</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"a43e7fba-a780-405c-ba92-f66d76274eeb\" type=\"checkbox\" ><label for=\"a43e7fba-a780-405c-ba92-f66d76274eeb\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ColumnSelectTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnSelectTransformer(columns=[&#x27;pclass&#x27;, &#x27;sex&#x27;, &#x27;age&#x27;, &#x27;sibsp&#x27;, &#x27;parch&#x27;,\n",
       "                                 &#x27;fare&#x27;, &#x27;embarked&#x27;])</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"3b4f7c77-a9f3-4d09-a10f-65dad22e1d2c\" type=\"checkbox\" ><label for=\"3b4f7c77-a9f3-4d09-a10f-65dad22e1d2c\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">LowerTransformer</label><div class=\"sk-toggleable__content\"><pre>LowerTransformer()</pre></div></div></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"870078a2-1b67-437e-a477-7be598acbc74\" type=\"checkbox\" ><label for=\"870078a2-1b67-437e-a477-7be598acbc74\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">label_encoding: columnTransformer</label><div class=\"sk-toggleable__content\"><pre>columnTransformer(remainder=&#x27;passthrough&#x27;,\n",
       "                  transformers=[(&#x27;category_encoder&#x27;,\n",
       "                                 CustomLabelEncoder(columns=[&#x27;surname&#x27;,\n",
       "                                                             &#x27;forename&#x27;,\n",
       "                                                             &#x27;ticket_str&#x27;,\n",
       "                                                             &#x27;cabin_str&#x27;, &#x27;sex&#x27;,\n",
       "                                                             &#x27;embarked&#x27;]),\n",
       "                                 [&#x27;surname&#x27;, &#x27;forename&#x27;, &#x27;ticket_str&#x27;,\n",
       "                                  &#x27;cabin_str&#x27;, &#x27;sex&#x27;, &#x27;embarked&#x27;])])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"8b93469b-1892-46a5-a3ec-9641082eb2f6\" type=\"checkbox\" ><label for=\"8b93469b-1892-46a5-a3ec-9641082eb2f6\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">category_encoder</label><div class=\"sk-toggleable__content\"><pre>[&#x27;surname&#x27;, &#x27;forename&#x27;, &#x27;ticket_str&#x27;, &#x27;cabin_str&#x27;, &#x27;sex&#x27;, &#x27;embarked&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"1c648523-dc96-4dcf-9a2f-69c0762e21a1\" type=\"checkbox\" ><label for=\"1c648523-dc96-4dcf-9a2f-69c0762e21a1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CustomLabelEncoder</label><div class=\"sk-toggleable__content\"><pre>CustomLabelEncoder(columns=[&#x27;surname&#x27;, &#x27;forename&#x27;, &#x27;ticket_str&#x27;, &#x27;cabin_str&#x27;,\n",
       "                            &#x27;sex&#x27;, &#x27;embarked&#x27;])</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"b2a41980-122e-4c0f-8ce0-f4db9928cbe8\" type=\"checkbox\" ><label for=\"b2a41980-122e-4c0f-8ce0-f4db9928cbe8\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">remainder</label><div class=\"sk-toggleable__content\"><pre>[&#x27;ticket_num&#x27;, &#x27;cabin_num&#x27;, &#x27;pclass&#x27;, &#x27;age&#x27;, &#x27;sibsp&#x27;, &#x27;parch&#x27;, &#x27;fare&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"de175232-9b36-4ba3-8f03-e212135052cc\" type=\"checkbox\" ><label for=\"de175232-9b36-4ba3-8f03-e212135052cc\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">passthrough</label><div class=\"sk-toggleable__content\"><pre>passthrough</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"c3b38134-1284-487c-ab24-2380d7897be8\" type=\"checkbox\" ><label for=\"c3b38134-1284-487c-ab24-2380d7897be8\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CustomLGBMClassifier</label><div class=\"sk-toggleable__content\"><pre>CustomLGBMClassifier()</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('prepro',\n",
       "                 Pipeline(steps=[('raw_data_processor',\n",
       "                                  featureUnion(transformer_list=[('pass_name',\n",
       "                                                                  PassNameTransformer()),\n",
       "                                                                 ('ticket',\n",
       "                                                                  TicketTransformer()),\n",
       "                                                                 ('cabin',\n",
       "                                                                  CabinTransformer()),\n",
       "                                                                 ('others',\n",
       "                                                                  ColumnSelectTransformer(columns=['pclass',\n",
       "                                                                                                   'sex',\n",
       "                                                                                                   'age',\n",
       "                                                                                                   'sibsp',\n",
       "                                                                                                   'parch',\n",
       "                                                                                                   'fare',\n",
       "                                                                                                   'embarked']))])),\n",
       "                                 ('lower_text_values', LowerTransformer())])),\n",
       "                ('label_encoding',\n",
       "                 columnTransformer(remainder='passthrough',\n",
       "                                   transformers=[('category_encoder',\n",
       "                                                  CustomLabelEncoder(columns=['surname',\n",
       "                                                                              'forename',\n",
       "                                                                              'ticket_str',\n",
       "                                                                              'cabin_str',\n",
       "                                                                              'sex',\n",
       "                                                                              'embarked']),\n",
       "                                                  ['surname', 'forename',\n",
       "                                                   'ticket_str', 'cabin_str',\n",
       "                                                   'sex', 'embarked'])])),\n",
       "                ('lgbm', CustomLGBMClassifier())])"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl_lgbm.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "52e76bd348784655baf173ee9bbe05ac",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[['surname', 'forename', 'ticket_str', 'cabin_str', 'sex', 'embarked'], ['ticket_num', 'cabin_num', 'pclass', 'age', 'sibsp', 'parch', 'fare']]\n",
      "[(100000, 6), (100000, 7)]\n",
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f37d36cb9f04438ebeefb7cd050174a8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "array([0, 0, 1, ..., 0, 1, 1])"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl_lgbm.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4b5bc38102d2413c9193270abd691ddb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[['surname', 'forename', 'ticket_str', 'cabin_str', 'sex', 'embarked'], ['ticket_num', 'cabin_num', 'pclass', 'age', 'sibsp', 'parch', 'fare']]\n",
      "[(100000, 6), (100000, 7)]\n",
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b36f35794a574d1bb65ab1e384041218",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "a = pl_lgbm.predict(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.81      0.83      0.82     57226\n",
      "           1       0.77      0.74      0.75     42774\n",
      "\n",
      "    accuracy                           0.79    100000\n",
      "   macro avg       0.79      0.79      0.79    100000\n",
      "weighted avg       0.79      0.79      0.79    100000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_train, a))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, ..., 0, 0, 0])"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b6e4b6bab0784514afe9b944c01b8123",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fdad906870e6463aa41ad6d92234c1a4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dc30c57a59f1415fbdd07d33d05f08b1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cb9d40f4a18042468d8b8d67362822b6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[['surname', 'forename', 'ticket_str', 'cabin_str', 'sex', 'embarked'], ['ticket_num', 'cabin_num', 'age', 'fare'], ['pclass', 'sibsp', 'parch']]\n",
      "[(100000, 6), (100000, 4), (100000, 3)]\n",
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "05e29109b3f84bd59b79762b5e71c68e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "b = pl4.predict(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, ..., 0, 0, 0])"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2451"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(~(a == b)).sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "from catboost import CatBoostRegressor, CatBoostClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fill_unk(sr):\n",
    "    return sr.fillna('unk')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fill_mean(sr):\n",
    "    return sr.fillna(sr.mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fill_0(sr):\n",
    "    return sr.fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fill_popular(sr):\n",
    "    popular = sr.value_counts().index[0]\n",
    "    return sr.fillna(popular)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'pc': 1.0,\n",
       " 'a.': 1.172558924272542,\n",
       " 'c.a.': 1.4930070292670683,\n",
       " 'ston/o': 1.5398503617138108,\n",
       " 'sc/paris': 1.546324951420759,\n",
       " 'a/5': 1.7613713129371242,\n",
       " 'soton/o.q.': 1.9118677832994102,\n",
       " 'w./c.': 1.9544197633354434,\n",
       " 'a/5.': 1.97821233213953,\n",
       " 'pp': 2.010050373472857,\n",
       " 'f.c.c.': 2.055849472309812,\n",
       " 'sc/ah': 2.1149313053493453,\n",
       " 'ca.': 2.156525142682023,\n",
       " 'ston/o2.': 2.2678590415219233,\n",
       " 'a/4': 2.317558829652617,\n",
       " 's.o.c.': 2.4891674024289627,\n",
       " 'f.c.': 2.511635098481276,\n",
       " 'soton/o2': 2.5271129738488165,\n",
       " 's.c./paris': 2.5350088959312904,\n",
       " 'ca': 2.5538574204092668,\n",
       " 'soton/oq': 2.6203790569660126,\n",
       " 'c': 2.7454734768560662,\n",
       " 'we/p': 2.9405775829198197,\n",
       " 'w.e.p.': 2.9715158105213897,\n",
       " 'sc': 2.9984092038990555,\n",
       " 'p/pp': 3.0859449625553044,\n",
       " 's.o./p.p.': 3.1857567885616125,\n",
       " 'a/4.': 3.2374490358210317,\n",
       " 'a./5.': 3.4178444779790533,\n",
       " 'sco/w': 3.467748915615144,\n",
       " 'a.5.': 3.776591219251863,\n",
       " 'aq/4': 3.8224974000984213,\n",
       " 'lp': 4.138925875399039,\n",
       " 'fa': 4.184330006248806,\n",
       " 'a/s': 4.449469701311798,\n",
       " 's.c./a.4.': 4.5785299490195,\n",
       " 'sw/pp': 4.649529181069397,\n",
       " 'ston/oq.': 4.649529181069397,\n",
       " 'c.a./soton': 4.686876819169658,\n",
       " 's.o.p.': 4.725560269497254,\n",
       " 'so/c': 4.807290662654624,\n",
       " 's.p.': 4.807290662654624,\n",
       " 'sc/a.3': 4.850537592242635,\n",
       " 's.w./pp': 4.942373343139944,\n",
       " 'sc/a4': 4.942373343139944,\n",
       " 'aq/3.': 4.991231118706202,\n",
       " 'w/c': 5.095615468746244,\n",
       " 'a4.': 5.724241875957752}"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "((df2['ticket_str'].value_counts().max() / df2['ticket_str'].value_counts()) ** 0.3).to_dict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'c': 1.0,\n",
       " 'b': 1.1491744912926358,\n",
       " 'a': 1.2075176724525294,\n",
       " 'd': 1.4243520812700914,\n",
       " 'e': 1.774203378786344,\n",
       " 'f': 2.373483879107566,\n",
       " 'g': 2.611717706060768,\n",
       " 't': 5.892442473806596}"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "((df2['cabin_str'].value_counts().max() / df2['cabin_str'].value_counts()) ** 0.3).to_dict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "impute_plan = {\n",
    "    'features_plan': {\n",
    "        'fare': {\n",
    "            'model': CatBoostRegressor,\n",
    "            'predictors': [c for c in df2.columns if c not in ['fare']],\n",
    "            'catboost_kwgs': {'iterations':200}\n",
    "        },\n",
    "        'embarked': {\n",
    "            'model': CatBoostClassifier,\n",
    "            'predictors': [c for c in df2.columns if c not in ['embarked']],\n",
    "            'catboost_kwgs': {'iterations': 200}\n",
    "        },\n",
    "        'age': {\n",
    "            'model': CatBoostRegressor,\n",
    "            'predictors': [c for c in df2.columns if c not in ['age']],\n",
    "            'catboost_kwgs': {'iterations': 200}\n",
    "        },\n",
    "        'ticket_num':{\n",
    "            'model': CatBoostRegressor,\n",
    "            'predictors': [c for c in df2.columns if c not in ['age']],\n",
    "            'catboost_kwgs': {'iterations': 200}\n",
    "        },\n",
    "        'ticket_str':{\n",
    "            'model': CatBoostClassifier,\n",
    "            'predictors': [c for c in df2.columns if c not in ['embarked']],\n",
    "            'catboost_kwgs': {\n",
    "                'iterations': 200, \n",
    "                'class_weights': ((df2['ticket_str'].value_counts().max() / df2['ticket_str'].value_counts()) ** 0.3).to_dict()\n",
    "            }\n",
    "        },\n",
    "        'cabin_num':{\n",
    "            'model': CatBoostRegressor,\n",
    "            'predictors': [c for c in df2.columns if c not in ['age']],\n",
    "            'catboost_kwgs': {\n",
    "                'iterations': 200\n",
    "            }\n",
    "        },\n",
    "        'cabin_str':{\n",
    "            'model': CatBoostClassifier,\n",
    "            'predictors': [c for c in df2.columns if c not in ['embarked']],\n",
    "            'catboost_kwgs': {\n",
    "                'iterations': 200,\n",
    "                'class_weights': ((df2['cabin_str'].value_counts().max() / df2['cabin_str'].value_counts()) ** 0.3).to_dict()\n",
    "            }\n",
    "        }\n",
    "    },\n",
    "    'if_na': {\n",
    "        'ticket_str':fill_unk,\n",
    "        'ticket_num':fill_mean,\n",
    "        'cabin_str':fill_unk,\n",
    "        'cabin_num':fill_mean,\n",
    "        'age':fill_mean,\n",
    "        'embarked':fill_popular,\n",
    "        'fare': fill_mean\n",
    "    },\n",
    "    'eval_frac': 0.1,\n",
    "    'cat_features': ['surname', 'forename', 'ticket_str', 'cabin_str', 'sex', 'embarked']\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "class CatBoostImputor(BaseEstimator, TransformerMixin):\n",
    "    def __init__(self, impute_plan, predictors_order):\n",
    "        super(CatBoostImputor).__init__()\n",
    "        self.impute_plan = impute_plan\n",
    "        self.predictors_order = predictors_order\n",
    "        \n",
    "    def fit(self, X, y=None):\n",
    "        return self\n",
    "    \n",
    "    def train_eval_test_split(self, x, y):\n",
    "        eval_frac = self.impute_plan['eval_frac']\n",
    "        y_test = y[y.isna()].copy()\n",
    "        x_test = x[y.isna()].copy()\n",
    "        x_t = x[~y.isna()].copy()\n",
    "        y_t = y[~y.isna()].copy()\n",
    "        x_train, x_eval, y_train, y_eval = train_test_split(x_t, y_t, test_size=eval_frac)\n",
    "        return x_train, y_train, x_eval, y_eval, x_test, y_test\n",
    "        \n",
    "    def transform(self, X):\n",
    "        self.X = X\n",
    "        self.X_impute = X.copy()\n",
    "        for col in self.predictors_order:\n",
    "            predictor_list = self.impute_plan['features_plan'][col]['predictors']\n",
    "            cat_features=[c for c in predictor_list if c in self.impute_plan['cat_features']]\n",
    "            # Prepare data to fit\n",
    "            print(f'Prepare data to fit, feature: {col}')\n",
    "            x = self.X_impute[predictor_list].copy()\n",
    "            y = self.X_impute[col]\n",
    "            na_cols = [col for col in x.columns if x[col].isna().sum() > 0]\n",
    "            for c in na_cols:\n",
    "                x[c] = self.impute_plan['if_na'][c](x[c])\n",
    "            x_train, y_train, x_eval, y_eval, x_test, y_test = self.train_eval_test_split(x, y)\n",
    "            \n",
    "            # Create model\n",
    "            ModelClass = self.impute_plan['features_plan'][col]['model']\n",
    "            model = ModelClass(**self.impute_plan['features_plan'][col]['catboost_kwgs'], early_stopping_rounds=50, verbose=True, metric_period=50, use_best_model=True)\n",
    "            # Fit model\n",
    "            print(f'Fit model: {col}')\n",
    "            model.fit(x_train, y_train, eval_set=(x_eval, y_eval), cat_features=cat_features)\n",
    "            x_test[col] = model.predict(x_test)\n",
    "            print(f'Predicted for {x_test.shape[0]} unknown examples')\n",
    "            self.X_impute.loc[y.isna(),col] = x_test[col]\n",
    "            print(f'Imputed done for {col}')\n",
    "        print(f'Imputed done for all dataset, num null left: {self.X_impute.isna().sum()}')\n",
    "        return self.X_impute\n",
    "            \n",
    "            "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "cbi = CatBoostImputor(impute_plan, ['fare', 'embarked', 'age', 'ticket_num', 'ticket_str', 'cabin_num', 'cabin_str'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prepare data to fit, feature: fare\n",
      "Fit model: fare\n",
      "Learning rate set to 0.275418\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:\tlearn: 64.5433575\ttest: 67.7969902\tbest: 67.7969902 (0)\ttotal: 31.6ms\tremaining: 6.3s\n",
      "50:\tlearn: 57.9453659\ttest: 61.5142867\tbest: 61.4951671 (21)\ttotal: 1.22s\tremaining: 3.57s\n",
      "Stopped by overfitting detector  (50 iterations wait)\n",
      "\n",
      "bestTest = 61.49516707\n",
      "bestIteration = 21\n",
      "\n",
      "Shrink model to first 22 iterations.\n",
      "Predicted for 134 unknown examples\n",
      "Imputed done for fare\n",
      "Prepare data to fit, feature: embarked\n",
      "Fit model: embarked\n",
      "Learning rate set to 0.218473\n",
      "0:\tlearn: 0.9030884\ttest: 0.9021783\tbest: 0.9021783 (0)\ttotal: 64ms\tremaining: 12.7s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 0.5340831\ttest: 0.5315682\tbest: 0.5315682 (50)\ttotal: 3.13s\tremaining: 9.15s\n",
      "100:\tlearn: 0.5255015\ttest: 0.5279925\tbest: 0.5278905 (98)\ttotal: 6.19s\tremaining: 6.07s\n",
      "150:\tlearn: 0.5191624\ttest: 0.5266327\tbest: 0.5266237 (148)\ttotal: 9.65s\tremaining: 3.13s\n",
      "199:\tlearn: 0.5141725\ttest: 0.5268364\tbest: 0.5265504 (160)\ttotal: 12.6s\tremaining: 0us\n",
      "\n",
      "bestTest = 0.5265504268\n",
      "bestIteration = 160\n",
      "\n",
      "Shrink model to first 161 iterations.\n",
      "Predicted for 250 unknown examples\n",
      "Imputed done for embarked\n",
      "Prepare data to fit, feature: age\n",
      "Fit model: age\n",
      "Learning rate set to 0.274032\n",
      "0:\tlearn: 17.5375367\ttest: 17.7804643\tbest: 17.7804643 (0)\ttotal: 32.6ms\tremaining: 6.49s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 16.3131702\ttest: 16.6063424\tbest: 16.6063424 (50)\ttotal: 1.12s\tremaining: 3.27s\n",
      "100:\tlearn: 16.2367186\ttest: 16.6074796\tbest: 16.6041532 (90)\ttotal: 2.39s\tremaining: 2.34s\n",
      "Stopped by overfitting detector  (50 iterations wait)\n",
      "\n",
      "bestTest = 16.60415325\n",
      "bestIteration = 90\n",
      "\n",
      "Shrink model to first 91 iterations.\n",
      "Predicted for 3292 unknown examples\n",
      "Imputed done for age\n",
      "Prepare data to fit, feature: ticket_num\n",
      "Fit model: ticket_num\n",
      "Learning rate set to 0.260893\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:\tlearn: 615484.1136638\ttest: 599467.5510643\tbest: 599467.5510643 (0)\ttotal: 11.8ms\tremaining: 2.34s\n",
      "50:\tlearn: 42526.2631765\ttest: 71569.0752631\tbest: 71141.7889173 (27)\ttotal: 831ms\tremaining: 2.43s\n",
      "Stopped by overfitting detector  (50 iterations wait)\n",
      "\n",
      "bestTest = 71141.78892\n",
      "bestIteration = 27\n",
      "\n",
      "Shrink model to first 28 iterations.\n",
      "Predicted for 29278 unknown examples\n",
      "Imputed done for ticket_num\n",
      "Prepare data to fit, feature: ticket_str\n",
      "Fit model: ticket_str\n",
      "Learning rate set to 0.21245\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:\tlearn: 1.9873531\ttest: 1.9674236\tbest: 1.9674236 (0)\ttotal: 2.59s\tremaining: 8m 34s\n",
      "50:\tlearn: 0.0970141\ttest: 0.0722092\tbest: 0.0722092 (50)\ttotal: 3m 6s\tremaining: 9m 4s\n",
      "100:\tlearn: 0.0586940\ttest: 0.0468815\tbest: 0.0468815 (100)\ttotal: 6m 22s\tremaining: 6m 14s\n",
      "150:\tlearn: 0.0382330\ttest: 0.0336705\tbest: 0.0336408 (149)\ttotal: 9m 53s\tremaining: 3m 12s\n",
      "199:\tlearn: 0.0272232\ttest: 0.0267822\tbest: 0.0267795 (198)\ttotal: 13m 26s\tremaining: 0us\n",
      "\n",
      "bestTest = 0.02677948147\n",
      "bestIteration = 198\n",
      "\n",
      "Shrink model to first 199 iterations.\n",
      "Predicted for 75345 unknown examples\n",
      "Imputed done for ticket_str\n",
      "Prepare data to fit, feature: cabin_num\n",
      "Fit model: cabin_num\n",
      "Learning rate set to 0.230503\n",
      "0:\tlearn: 4502.7071399\ttest: 4486.3479850\tbest: 4486.3479850 (0)\ttotal: 26.2ms\tremaining: 5.22s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 120.5011194\ttest: 120.1827804\tbest: 120.1827804 (50)\ttotal: 576ms\tremaining: 1.68s\n",
      "100:\tlearn: 70.6091375\ttest: 70.4021427\tbest: 70.4021427 (100)\ttotal: 1.19s\tremaining: 1.17s\n",
      "150:\tlearn: 53.0992116\ttest: 53.2498008\tbest: 53.2498008 (150)\ttotal: 1.75s\tremaining: 566ms\n",
      "199:\tlearn: 44.2647913\ttest: 44.5245671\tbest: 44.5181596 (198)\ttotal: 2.21s\tremaining: 0us\n",
      "\n",
      "bestTest = 44.51815961\n",
      "bestIteration = 198\n",
      "\n",
      "Shrink model to first 199 iterations.\n",
      "Predicted for 67866 unknown examples\n",
      "Imputed done for cabin_num\n",
      "Prepare data to fit, feature: cabin_str\n",
      "Fit model: cabin_str\n",
      "Learning rate set to 0.213579\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:\tlearn: 0.8352189\ttest: 0.8447905\tbest: 0.8447905 (0)\ttotal: 129ms\tremaining: 25.6s\n",
      "50:\tlearn: 0.0019150\ttest: 0.0015528\tbest: 0.0015528 (50)\ttotal: 5.26s\tremaining: 15.4s\n",
      "100:\tlearn: 0.0007051\ttest: 0.0005527\tbest: 0.0005527 (100)\ttotal: 11.1s\tremaining: 10.9s\n",
      "150:\tlearn: 0.0003813\ttest: 0.0002886\tbest: 0.0002886 (150)\ttotal: 16.8s\tremaining: 5.46s\n",
      "199:\tlearn: 0.0002591\ttest: 0.0001920\tbest: 0.0001920 (199)\ttotal: 23.1s\tremaining: 0us\n",
      "\n",
      "bestTest = 0.0001920258865\n",
      "bestIteration = 199\n",
      "\n",
      "Predicted for 67866 unknown examples\n",
      "Imputed done for cabin_str\n",
      "Imputed done for all dataset, num null left: surname       0\n",
      "forename      0\n",
      "ticket_str    0\n",
      "ticket_num    0\n",
      "cabin_str     0\n",
      "cabin_num     0\n",
      "pclass        0\n",
      "sex           0\n",
      "age           0\n",
      "sibsp         0\n",
      "parch         0\n",
      "fare          0\n",
      "embarked      0\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "clean_X_train = cbi.fit_transform(df2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Done\n"
     ]
    }
   ],
   "source": [
    "print('Done')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "clean_X_train.to_parquet('data/clean_X_train.parquet')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "passengerid\n",
       "0        1\n",
       "1        0\n",
       "2        0\n",
       "3        0\n",
       "4        1\n",
       "        ..\n",
       "99995    1\n",
       "99996    0\n",
       "99997    0\n",
       "99998    0\n",
       "99999    0\n",
       "Name: survived, Length: 100000, dtype: int64"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_t, X_train_e, y_train_t, y_train_e = train_test_split(clean_X_train, y_train, test_size=0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_clean = CatBoostClassifier(iterations=1000, cat_features=['surname', 'forename', 'ticket_str', 'cabin_str', 'sex', 'embarked'], \n",
    "                                 early_stopping_rounds=50, verbose=True, metric_period=50, use_best_model=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>surname</th>\n",
       "      <th>forename</th>\n",
       "      <th>ticket_str</th>\n",
       "      <th>ticket_num</th>\n",
       "      <th>cabin_str</th>\n",
       "      <th>cabin_num</th>\n",
       "      <th>pclass</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>sibsp</th>\n",
       "      <th>parch</th>\n",
       "      <th>fare</th>\n",
       "      <th>embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>48829</th>\n",
       "      <td>benefield</td>\n",
       "      <td>mary</td>\n",
       "      <td>fa</td>\n",
       "      <td>459547.000000</td>\n",
       "      <td>t</td>\n",
       "      <td>12541.295461</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>42.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.42</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51905</th>\n",
       "      <td>kirkpatrick</td>\n",
       "      <td>bertha</td>\n",
       "      <td>s.o.p.</td>\n",
       "      <td>480132.000000</td>\n",
       "      <td>t</td>\n",
       "      <td>12529.354019</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>84.42</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74339</th>\n",
       "      <td>knight</td>\n",
       "      <td>pete</td>\n",
       "      <td>so/c</td>\n",
       "      <td>30991.000000</td>\n",
       "      <td>t</td>\n",
       "      <td>12513.448282</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>29.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>26.42</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73410</th>\n",
       "      <td>mottillo</td>\n",
       "      <td>lucille</td>\n",
       "      <td>a4.</td>\n",
       "      <td>400759.000000</td>\n",
       "      <td>t</td>\n",
       "      <td>12542.585403</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>55.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10.89</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66487</th>\n",
       "      <td>gomez</td>\n",
       "      <td>kimberly</td>\n",
       "      <td>w/c</td>\n",
       "      <td>25378.000000</td>\n",
       "      <td>b</td>\n",
       "      <td>19913.000000</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>62.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>20.52</td>\n",
       "      <td>q</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66214</th>\n",
       "      <td>feldmann</td>\n",
       "      <td>matthew</td>\n",
       "      <td>so/c</td>\n",
       "      <td>195108.000000</td>\n",
       "      <td>t</td>\n",
       "      <td>12513.495526</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>25.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>9.88</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81785</th>\n",
       "      <td>vroman</td>\n",
       "      <td>patsy</td>\n",
       "      <td>ston/o2.</td>\n",
       "      <td>345541.968025</td>\n",
       "      <td>t</td>\n",
       "      <td>12541.671085</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>9.70</td>\n",
       "      <td>c</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4888</th>\n",
       "      <td>hayes</td>\n",
       "      <td>vernon</td>\n",
       "      <td>a/s</td>\n",
       "      <td>211552.000000</td>\n",
       "      <td>t</td>\n",
       "      <td>12513.845691</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>25.0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>27.83</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5291</th>\n",
       "      <td>thompson</td>\n",
       "      <td>michael</td>\n",
       "      <td>so/c</td>\n",
       "      <td>40641.000000</td>\n",
       "      <td>t</td>\n",
       "      <td>12513.495526</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>36.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>23.35</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9105</th>\n",
       "      <td>jackson</td>\n",
       "      <td>janel</td>\n",
       "      <td>sc/a4</td>\n",
       "      <td>16900.000000</td>\n",
       "      <td>t</td>\n",
       "      <td>12513.763400</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>74.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10.79</td>\n",
       "      <td>c</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>90000 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           surname   forename ticket_str     ticket_num cabin_str  \\\n",
       "48829    benefield       mary         fa  459547.000000         t   \n",
       "51905  kirkpatrick     bertha     s.o.p.  480132.000000         t   \n",
       "74339       knight       pete       so/c   30991.000000         t   \n",
       "73410     mottillo    lucille        a4.  400759.000000         t   \n",
       "66487        gomez   kimberly        w/c   25378.000000         b   \n",
       "...            ...        ...        ...            ...       ...   \n",
       "66214     feldmann    matthew       so/c  195108.000000         t   \n",
       "81785       vroman      patsy   ston/o2.  345541.968025         t   \n",
       "4888         hayes     vernon        a/s  211552.000000         t   \n",
       "5291      thompson    michael       so/c   40641.000000         t   \n",
       "9105       jackson      janel      sc/a4   16900.000000         t   \n",
       "\n",
       "          cabin_num  pclass     sex   age  sibsp  parch   fare embarked  \n",
       "48829  12541.295461       3  female  42.0      0      0   5.42        s  \n",
       "51905  12529.354019       3  female   7.0      1      2  84.42        s  \n",
       "74339  12513.448282       3    male  29.0      1      0  26.42        s  \n",
       "73410  12542.585403       3  female  55.0      0      0  10.89        s  \n",
       "66487  19913.000000       1  female  62.0      0      0  20.52        q  \n",
       "...             ...     ...     ...   ...    ...    ...    ...      ...  \n",
       "66214  12513.495526       3    male  25.0      0      0   9.88        s  \n",
       "81785  12541.671085       3  female   9.0      0      2   9.70        c  \n",
       "4888   12513.845691       3    male  25.0      0      3  27.83        s  \n",
       "5291   12513.495526       3    male  36.0      0      0  23.35        s  \n",
       "9105   12513.763400       3  female  74.0      0      0  10.79        c  \n",
       "\n",
       "[90000 rows x 13 columns]"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_t"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Learning rate set to 0.096296\n",
      "0:\tlearn: 0.6506654\ttest: 0.6502119\tbest: 0.6502119 (0)\ttotal: 42.5ms\tremaining: 42.5s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 0.4696837\ttest: 0.4695765\tbest: 0.4695765 (50)\ttotal: 1.62s\tremaining: 30.2s\n",
      "100:\tlearn: 0.4640451\ttest: 0.4666073\tbest: 0.4666073 (100)\ttotal: 3.33s\tremaining: 29.6s\n",
      "150:\tlearn: 0.4604643\ttest: 0.4654303\tbest: 0.4654274 (146)\ttotal: 5.4s\tremaining: 30.4s\n",
      "200:\tlearn: 0.4575480\ttest: 0.4651951\tbest: 0.4651719 (199)\ttotal: 7.26s\tremaining: 28.8s\n",
      "250:\tlearn: 0.4548392\ttest: 0.4649142\tbest: 0.4649071 (249)\ttotal: 9.06s\tremaining: 27s\n",
      "300:\tlearn: 0.4523819\ttest: 0.4647689\tbest: 0.4646811 (287)\ttotal: 10.8s\tremaining: 25s\n",
      "350:\tlearn: 0.4502126\ttest: 0.4644886\tbest: 0.4644594 (341)\ttotal: 12.5s\tremaining: 23.1s\n",
      "400:\tlearn: 0.4481511\ttest: 0.4643741\tbest: 0.4643461 (397)\ttotal: 14.3s\tremaining: 21.3s\n",
      "450:\tlearn: 0.4462194\ttest: 0.4644705\tbest: 0.4643258 (404)\ttotal: 15.9s\tremaining: 19.3s\n",
      "Stopped by overfitting detector  (50 iterations wait)\n",
      "\n",
      "bestTest = 0.4643258221\n",
      "bestIteration = 404\n",
      "\n",
      "Shrink model to first 405 iterations.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<catboost.core.CatBoostClassifier at 0x7f84b91361f0>"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_clean.fit(X_train_t, y_train_t, eval_set = (X_train_e, y_train_e))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train_e_pred = model_clean.predict(X_train_e)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.81      0.82      0.82      5744\n",
      "           1       0.75      0.74      0.75      4256\n",
      "\n",
      "    accuracy                           0.79     10000\n",
      "   macro avg       0.78      0.78      0.78     10000\n",
      "weighted avg       0.79      0.79      0.79     10000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "print(classification_report(y_train_e, model_clean.predict(X_train_e)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.81      0.83      0.82     51482\n",
      "           1       0.76      0.74      0.75     38518\n",
      "\n",
      "    accuracy                           0.79     90000\n",
      "   macro avg       0.79      0.78      0.78     90000\n",
      "weighted avg       0.79      0.79      0.79     90000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_train_t, model_clean.predict(X_train_t)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "def ft_imp(model):\n",
    "    fi = pd.DataFrame({\n",
    "        'feature':X_train_e.columns,\n",
    "        'importance': model.feature_importances_\n",
    "    })\n",
    "\n",
    "    return fi.sort_values(by='importance',ascending=False).iloc[:50].style.bar()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "#T_cfb90_row0_col1{\n",
       "            width:  10em;\n",
       "             height:  80%;\n",
       "            background:  linear-gradient(90deg,#d65f5f 100.0%, transparent 100.0%);\n",
       "        }#T_cfb90_row1_col1{\n",
       "            width:  10em;\n",
       "             height:  80%;\n",
       "            background:  linear-gradient(90deg,#d65f5f 40.8%, transparent 40.8%);\n",
       "        }#T_cfb90_row2_col1{\n",
       "            width:  10em;\n",
       "             height:  80%;\n",
       "            background:  linear-gradient(90deg,#d65f5f 37.9%, transparent 37.9%);\n",
       "        }#T_cfb90_row3_col1{\n",
       "            width:  10em;\n",
       "             height:  80%;\n",
       "            background:  linear-gradient(90deg,#d65f5f 37.0%, transparent 37.0%);\n",
       "        }#T_cfb90_row4_col1{\n",
       "            width:  10em;\n",
       "             height:  80%;\n",
       "            background:  linear-gradient(90deg,#d65f5f 31.3%, transparent 31.3%);\n",
       "        }#T_cfb90_row5_col1{\n",
       "            width:  10em;\n",
       "             height:  80%;\n",
       "            background:  linear-gradient(90deg,#d65f5f 31.1%, transparent 31.1%);\n",
       "        }#T_cfb90_row6_col1{\n",
       "            width:  10em;\n",
       "             height:  80%;\n",
       "            background:  linear-gradient(90deg,#d65f5f 28.9%, transparent 28.9%);\n",
       "        }#T_cfb90_row7_col1{\n",
       "            width:  10em;\n",
       "             height:  80%;\n",
       "            background:  linear-gradient(90deg,#d65f5f 14.1%, transparent 14.1%);\n",
       "        }#T_cfb90_row8_col1{\n",
       "            width:  10em;\n",
       "             height:  80%;\n",
       "            background:  linear-gradient(90deg,#d65f5f 13.0%, transparent 13.0%);\n",
       "        }#T_cfb90_row9_col1{\n",
       "            width:  10em;\n",
       "             height:  80%;\n",
       "            background:  linear-gradient(90deg,#d65f5f 9.4%, transparent 9.4%);\n",
       "        }#T_cfb90_row10_col1{\n",
       "            width:  10em;\n",
       "             height:  80%;\n",
       "            background:  linear-gradient(90deg,#d65f5f 8.0%, transparent 8.0%);\n",
       "        }#T_cfb90_row11_col1{\n",
       "            width:  10em;\n",
       "             height:  80%;\n",
       "            background:  linear-gradient(90deg,#d65f5f 7.2%, transparent 7.2%);\n",
       "        }#T_cfb90_row12_col1{\n",
       "            width:  10em;\n",
       "             height:  80%;\n",
       "        }</style><table id=\"T_cfb90_\" ><thead>    <tr>        <th class=\"blank level0\" ></th>        <th class=\"col_heading level0 col0\" >feature</th>        <th class=\"col_heading level0 col1\" >importance</th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_cfb90_level0_row0\" class=\"row_heading level0 row0\" >7</th>\n",
       "                        <td id=\"T_cfb90_row0_col0\" class=\"data row0 col0\" >sex</td>\n",
       "                        <td id=\"T_cfb90_row0_col1\" class=\"data row0 col1\" >23.536167</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_cfb90_level0_row1\" class=\"row_heading level0 row1\" >6</th>\n",
       "                        <td id=\"T_cfb90_row1_col0\" class=\"data row1 col0\" >pclass</td>\n",
       "                        <td id=\"T_cfb90_row1_col1\" class=\"data row1 col1\" >10.583626</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_cfb90_level0_row2\" class=\"row_heading level0 row2\" >11</th>\n",
       "                        <td id=\"T_cfb90_row2_col0\" class=\"data row2 col0\" >fare</td>\n",
       "                        <td id=\"T_cfb90_row2_col1\" class=\"data row2 col1\" >9.940771</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_cfb90_level0_row3\" class=\"row_heading level0 row3\" >12</th>\n",
       "                        <td id=\"T_cfb90_row3_col0\" class=\"data row3 col0\" >embarked</td>\n",
       "                        <td id=\"T_cfb90_row3_col1\" class=\"data row3 col1\" >9.754716</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_cfb90_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "                        <td id=\"T_cfb90_row4_col0\" class=\"data row4 col0\" >cabin_str</td>\n",
       "                        <td id=\"T_cfb90_row4_col1\" class=\"data row4 col1\" >8.509579</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_cfb90_level0_row5\" class=\"row_heading level0 row5\" >2</th>\n",
       "                        <td id=\"T_cfb90_row5_col0\" class=\"data row5 col0\" >ticket_str</td>\n",
       "                        <td id=\"T_cfb90_row5_col1\" class=\"data row5 col1\" >8.451626</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_cfb90_level0_row6\" class=\"row_heading level0 row6\" >3</th>\n",
       "                        <td id=\"T_cfb90_row6_col0\" class=\"data row6 col0\" >ticket_num</td>\n",
       "                        <td id=\"T_cfb90_row6_col1\" class=\"data row6 col1\" >7.974061</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_cfb90_level0_row7\" class=\"row_heading level0 row7\" >10</th>\n",
       "                        <td id=\"T_cfb90_row7_col0\" class=\"data row7 col0\" >parch</td>\n",
       "                        <td id=\"T_cfb90_row7_col1\" class=\"data row7 col1\" >4.732307</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_cfb90_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "                        <td id=\"T_cfb90_row8_col0\" class=\"data row8 col0\" >age</td>\n",
       "                        <td id=\"T_cfb90_row8_col1\" class=\"data row8 col1\" >4.507054</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_cfb90_level0_row9\" class=\"row_heading level0 row9\" >1</th>\n",
       "                        <td id=\"T_cfb90_row9_col0\" class=\"data row9 col0\" >forename</td>\n",
       "                        <td id=\"T_cfb90_row9_col1\" class=\"data row9 col1\" >3.717112</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_cfb90_level0_row10\" class=\"row_heading level0 row10\" >5</th>\n",
       "                        <td id=\"T_cfb90_row10_col0\" class=\"data row10 col0\" >cabin_num</td>\n",
       "                        <td id=\"T_cfb90_row10_col1\" class=\"data row10 col1\" >3.399783</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_cfb90_level0_row11\" class=\"row_heading level0 row11\" >9</th>\n",
       "                        <td id=\"T_cfb90_row11_col0\" class=\"data row11 col0\" >sibsp</td>\n",
       "                        <td id=\"T_cfb90_row11_col1\" class=\"data row11 col1\" >3.234756</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_cfb90_level0_row12\" class=\"row_heading level0 row12\" >0</th>\n",
       "                        <td id=\"T_cfb90_row12_col0\" class=\"data row12 col0\" >surname</td>\n",
       "                        <td id=\"T_cfb90_row12_col1\" class=\"data row12 col1\" >1.658442</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7f84d8ce00d0>"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ft_imp(model_clean)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9fe82356ed6c4c0fa29531028dc6c0e2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_test_2 = pl.fit_transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>surname</th>\n",
       "      <th>forename</th>\n",
       "      <th>ticket_str</th>\n",
       "      <th>ticket_num</th>\n",
       "      <th>cabin_str</th>\n",
       "      <th>cabin_num</th>\n",
       "      <th>pclass</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>sibsp</th>\n",
       "      <th>parch</th>\n",
       "      <th>fare</th>\n",
       "      <th>embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>holliday</td>\n",
       "      <td>daniel</td>\n",
       "      <td>None</td>\n",
       "      <td>24745.0</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>63.01</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>nguyen</td>\n",
       "      <td>lorraine</td>\n",
       "      <td>None</td>\n",
       "      <td>13264.0</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>53.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.81</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>harris</td>\n",
       "      <td>heather</td>\n",
       "      <td>None</td>\n",
       "      <td>25990.0</td>\n",
       "      <td>b</td>\n",
       "      <td>15315.0</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>38.91</td>\n",
       "      <td>c</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>larsen</td>\n",
       "      <td>eric</td>\n",
       "      <td>None</td>\n",
       "      <td>314011.0</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>male</td>\n",
       "      <td>25.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>12.93</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>cleary</td>\n",
       "      <td>sarah</td>\n",
       "      <td>None</td>\n",
       "      <td>26203.0</td>\n",
       "      <td>b</td>\n",
       "      <td>22515.0</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>17.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>26.89</td>\n",
       "      <td>c</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99995</th>\n",
       "      <td>cash</td>\n",
       "      <td>cheryle</td>\n",
       "      <td>None</td>\n",
       "      <td>7686.0</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10.12</td>\n",
       "      <td>q</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99996</th>\n",
       "      <td>brown</td>\n",
       "      <td>howard</td>\n",
       "      <td>None</td>\n",
       "      <td>13004.0</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>59.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>68.31</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99997</th>\n",
       "      <td>lightfoot</td>\n",
       "      <td>cameron</td>\n",
       "      <td>None</td>\n",
       "      <td>4383317.0</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>47.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10.87</td>\n",
       "      <td>s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99998</th>\n",
       "      <td>jacobsen</td>\n",
       "      <td>margaret</td>\n",
       "      <td>pc</td>\n",
       "      <td>NaN</td>\n",
       "      <td>b</td>\n",
       "      <td>20828.0</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>49.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>29.68</td>\n",
       "      <td>c</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99999</th>\n",
       "      <td>fishback</td>\n",
       "      <td>joanna</td>\n",
       "      <td>pc</td>\n",
       "      <td>NaN</td>\n",
       "      <td>e</td>\n",
       "      <td>13345.0</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>41.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>195.41</td>\n",
       "      <td>c</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100000 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         surname   forename ticket_str  ticket_num cabin_str  cabin_num  \\\n",
       "0       holliday     daniel       None     24745.0      None        NaN   \n",
       "1         nguyen   lorraine       None     13264.0      None        NaN   \n",
       "2         harris    heather       None     25990.0         b    15315.0   \n",
       "3         larsen       eric       None    314011.0      None        NaN   \n",
       "4         cleary      sarah       None     26203.0         b    22515.0   \n",
       "...          ...        ...        ...         ...       ...        ...   \n",
       "99995       cash    cheryle       None      7686.0      None        NaN   \n",
       "99996      brown     howard       None     13004.0      None        NaN   \n",
       "99997  lightfoot    cameron       None   4383317.0      None        NaN   \n",
       "99998   jacobsen   margaret         pc         NaN         b    20828.0   \n",
       "99999   fishback     joanna         pc         NaN         e    13345.0   \n",
       "\n",
       "       pclass     sex   age  sibsp  parch    fare embarked  \n",
       "0           3    male  19.0      0      0   63.01        s  \n",
       "1           3  female  53.0      0      0    5.81        s  \n",
       "2           1  female  19.0      0      0   38.91        c  \n",
       "3           2    male  25.0      0      0   12.93        s  \n",
       "4           1  female  17.0      0      2   26.89        c  \n",
       "...       ...     ...   ...    ...    ...     ...      ...  \n",
       "99995       3  female  27.0      0      0   10.12        q  \n",
       "99996       1    male  59.0      1      0   68.31        s  \n",
       "99997       3    male  47.0      0      0   10.87        s  \n",
       "99998       1  female  49.0      1      2   29.68        c  \n",
       "99999       1  female  41.0      0      2  195.41        c  \n",
       "\n",
       "[100000 rows x 13 columns]"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test_2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "cbi_test = CatBoostImputor(impute_plan, ['fare', 'embarked', 'age', 'ticket_num', 'ticket_str', 'cabin_num', 'cabin_str'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prepare data to fit, feature: fare\n",
      "Fit model: fare\n",
      "Learning rate set to 0.275418\n",
      "0:\tlearn: 55.4252142\ttest: 57.5502950\tbest: 57.5502950 (0)\ttotal: 23.2ms\tremaining: 4.62s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 41.0859880\ttest: 44.1781527\tbest: 44.1239668 (17)\ttotal: 1.36s\tremaining: 3.96s\n",
      "Stopped by overfitting detector  (50 iterations wait)\n",
      "\n",
      "bestTest = 44.12396684\n",
      "bestIteration = 17\n",
      "\n",
      "Shrink model to first 18 iterations.\n",
      "Predicted for 133 unknown examples\n",
      "Imputed done for fare\n",
      "Prepare data to fit, feature: embarked\n",
      "Fit model: embarked\n",
      "Learning rate set to 0.218472\n",
      "0:\tlearn: 0.9434240\ttest: 0.9419511\tbest: 0.9419511 (0)\ttotal: 65ms\tremaining: 12.9s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 0.6375239\ttest: 0.6444090\tbest: 0.6444090 (50)\ttotal: 3.19s\tremaining: 9.32s\n",
      "100:\tlearn: 0.6256712\ttest: 0.6384090\tbest: 0.6384090 (100)\ttotal: 6.29s\tremaining: 6.17s\n",
      "150:\tlearn: 0.6195778\ttest: 0.6376685\tbest: 0.6375410 (140)\ttotal: 9.34s\tremaining: 3.03s\n",
      "199:\tlearn: 0.6147086\ttest: 0.6377627\tbest: 0.6374074 (178)\ttotal: 12.3s\tremaining: 0us\n",
      "\n",
      "bestTest = 0.6374074293\n",
      "bestIteration = 178\n",
      "\n",
      "Shrink model to first 179 iterations.\n",
      "Predicted for 277 unknown examples\n",
      "Imputed done for embarked\n",
      "Prepare data to fit, feature: age\n",
      "Fit model: age\n",
      "Learning rate set to 0.273945\n",
      "0:\tlearn: 13.5703276\ttest: 13.5648230\tbest: 13.5648230 (0)\ttotal: 27.3ms\tremaining: 5.44s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 12.7572813\ttest: 12.8397431\tbest: 12.8395943 (47)\ttotal: 1.12s\tremaining: 3.28s\n",
      "100:\tlearn: 12.6899986\ttest: 12.8322579\tbest: 12.8316945 (91)\ttotal: 2.29s\tremaining: 2.25s\n",
      "150:\tlearn: 12.6354794\ttest: 12.8383657\tbest: 12.8308877 (102)\ttotal: 3.59s\tremaining: 1.17s\n",
      "Stopped by overfitting detector  (50 iterations wait)\n",
      "\n",
      "bestTest = 12.83088768\n",
      "bestIteration = 102\n",
      "\n",
      "Shrink model to first 103 iterations.\n",
      "Predicted for 3487 unknown examples\n",
      "Imputed done for age\n",
      "Prepare data to fit, feature: ticket_num\n",
      "Fit model: ticket_num\n",
      "Learning rate set to 0.260343\n",
      "0:\tlearn: 655725.7709749\ttest: 683463.2397403\tbest: 683463.2397403 (0)\ttotal: 11.7ms\tremaining: 2.33s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 89749.4989781\ttest: 98185.2575162\tbest: 95505.9245302 (28)\ttotal: 897ms\tremaining: 2.62s\n",
      "Stopped by overfitting detector  (50 iterations wait)\n",
      "\n",
      "bestTest = 95505.92453\n",
      "bestIteration = 28\n",
      "\n",
      "Shrink model to first 29 iterations.\n",
      "Predicted for 30223 unknown examples\n",
      "Imputed done for ticket_num\n",
      "Prepare data to fit, feature: ticket_str\n",
      "Fit model: ticket_str\n",
      "Learning rate set to 0.212516\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:\tlearn: 1.3240162\ttest: 1.3590509\tbest: 1.3590509 (0)\ttotal: 3s\tremaining: 9m 57s\n",
      "50:\tlearn: 0.1033778\ttest: 0.1039747\tbest: 0.1039747 (50)\ttotal: 3m 17s\tremaining: 9m 36s\n",
      "100:\tlearn: 0.0538585\ttest: 0.0677874\tbest: 0.0677392 (97)\ttotal: 6m 51s\tremaining: 6m 43s\n",
      "150:\tlearn: 0.0341126\ttest: 0.0528243\tbest: 0.0527686 (149)\ttotal: 10m 15s\tremaining: 3m 19s\n",
      "199:\tlearn: 0.0215465\ttest: 0.0421844\tbest: 0.0421844 (199)\ttotal: 14m 15s\tremaining: 0us\n",
      "\n",
      "bestTest = 0.04218435661\n",
      "bestIteration = 199\n",
      "\n",
      "Predicted for 74958 unknown examples\n",
      "Imputed done for ticket_str\n",
      "Prepare data to fit, feature: cabin_num\n",
      "Fit model: cabin_num\n",
      "Learning rate set to 0.227026\n",
      "0:\tlearn: 4413.0014249\ttest: 4498.4735967\tbest: 4498.4735967 (0)\ttotal: 67.4ms\tremaining: 13.4s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 122.8437567\ttest: 124.1164012\tbest: 124.1164012 (50)\ttotal: 766ms\tremaining: 2.24s\n",
      "100:\tlearn: 70.6002670\ttest: 72.8305021\tbest: 72.8305021 (100)\ttotal: 1.23s\tremaining: 1.21s\n",
      "150:\tlearn: 51.8690868\ttest: 54.5752248\tbest: 54.5752248 (150)\ttotal: 1.69s\tremaining: 548ms\n",
      "199:\tlearn: 43.5400815\ttest: 47.2411642\tbest: 47.2411642 (199)\ttotal: 2.19s\tremaining: 0us\n",
      "\n",
      "bestTest = 47.24116422\n",
      "bestIteration = 199\n",
      "\n",
      "Predicted for 70831 unknown examples\n",
      "Imputed done for cabin_num\n",
      "Prepare data to fit, feature: cabin_str\n",
      "Fit model: cabin_str\n",
      "Learning rate set to 0.213166\n",
      "0:\tlearn: 0.8367792\ttest: 0.8313201\tbest: 0.8313201 (0)\ttotal: 82.5ms\tremaining: 16.4s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 0.0016538\ttest: 0.0012683\tbest: 0.0012683 (50)\ttotal: 5.12s\tremaining: 15s\n",
      "100:\tlearn: 0.0006863\ttest: 0.0005304\tbest: 0.0005304 (100)\ttotal: 10.8s\tremaining: 10.6s\n",
      "150:\tlearn: 0.0004633\ttest: 0.0003657\tbest: 0.0003657 (150)\ttotal: 16.4s\tremaining: 5.33s\n",
      "199:\tlearn: 0.0003362\ttest: 0.0002622\tbest: 0.0002622 (199)\ttotal: 21.8s\tremaining: 0us\n",
      "\n",
      "bestTest = 0.0002621691329\n",
      "bestIteration = 199\n",
      "\n",
      "Predicted for 70831 unknown examples\n",
      "Imputed done for cabin_str\n",
      "Imputed done for all dataset, num null left: surname       0\n",
      "forename      0\n",
      "ticket_str    0\n",
      "ticket_num    0\n",
      "cabin_str     0\n",
      "cabin_num     0\n",
      "pclass        0\n",
      "sex           0\n",
      "age           0\n",
      "sibsp         0\n",
      "parch         0\n",
      "fare          0\n",
      "embarked      0\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "X_test_clean = cbi_test.fit_transform(X_test_2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test_clean.to_parquet('data/clean_X_test.parquet')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_test_2.ticket_num.hist(bins=50)\n",
    "plt.show()\n",
    "X_test_clean.ticket_num.hist(bins=50)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_test_2.cabin_num.hist(bins=50)\n",
    "plt.show()\n",
    "X_test_clean.cabin_num.hist(bins=50)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_test_2.age.hist(bins=50)\n",
    "plt.show()\n",
    "X_test_clean.age.hist(bins=50)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "b    8113\n",
       "a    7212\n",
       "c    6993\n",
       "d    2521\n",
       "f    2323\n",
       "e    1837\n",
       "g     144\n",
       "t      26\n",
       "Name: cabin_str, dtype: int64"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test_2.cabin_str.value_counts()[:20]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "t    70331\n",
       "b     8113\n",
       "a     7212\n",
       "c     6993\n",
       "d     2521\n",
       "f     2323\n",
       "e     1837\n",
       "g      670\n",
       "Name: cabin_str, dtype: int64"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test_clean.cabin_str.value_counts()[:20]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Extent pipeline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 {color: black;background-color: white;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 pre{padding: 0;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-toggleable {background-color: white;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-estimator:hover {background-color: #d4ebff;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-item {z-index: 1;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-parallel::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-parallel-item:only-child::after {width: 0;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;position: relative;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-label-container {position: relative;z-index: 2;text-align: center;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-a3671ece-6e73-4fba-a05a-f3126eaf4e08\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                 featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                 PassNameTransformer()),\n",
       "                                                (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                                                (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                                                (&#x27;others&#x27;,\n",
       "                                                 ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                  &#x27;sex&#x27;,\n",
       "                                                                                  &#x27;age&#x27;,\n",
       "                                                                                  &#x27;sibsp&#x27;,\n",
       "                                                                                  &#x27;parch&#x27;,\n",
       "                                                                                  &#x27;fare&#x27;,\n",
       "                                                                                  &#x27;embarked&#x27;]))])),\n",
       "                (&#x27;lower_text_values&#x27;, LowerTransformer())])</pre><b>Please rerun this cell to show the HTML repr or trust the notebook.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"c35e346e-7702-47b6-baec-e568ae26ff62\" type=\"checkbox\" ><label for=\"c35e346e-7702-47b6-baec-e568ae26ff62\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                 featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                 PassNameTransformer()),\n",
       "                                                (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                                                (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                                                (&#x27;others&#x27;,\n",
       "                                                 ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                  &#x27;sex&#x27;,\n",
       "                                                                                  &#x27;age&#x27;,\n",
       "                                                                                  &#x27;sibsp&#x27;,\n",
       "                                                                                  &#x27;parch&#x27;,\n",
       "                                                                                  &#x27;fare&#x27;,\n",
       "                                                                                  &#x27;embarked&#x27;]))])),\n",
       "                (&#x27;lower_text_values&#x27;, LowerTransformer())])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"8fc8e9bc-1a9c-4128-927e-1c3d8f40cd7b\" type=\"checkbox\" ><label for=\"8fc8e9bc-1a9c-4128-927e-1c3d8f40cd7b\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">raw_data_processor: featureUnion</label><div class=\"sk-toggleable__content\"><pre>featureUnion(transformer_list=[(&#x27;pass_name&#x27;, PassNameTransformer()),\n",
       "                               (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                               (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                               (&#x27;others&#x27;,\n",
       "                                ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                 &#x27;sex&#x27;, &#x27;age&#x27;,\n",
       "                                                                 &#x27;sibsp&#x27;,\n",
       "                                                                 &#x27;parch&#x27;,\n",
       "                                                                 &#x27;fare&#x27;,\n",
       "                                                                 &#x27;embarked&#x27;]))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>pass_name</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"e79eb053-3936-4b28-9db2-bb10c885dc00\" type=\"checkbox\" ><label for=\"e79eb053-3936-4b28-9db2-bb10c885dc00\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">PassNameTransformer</label><div class=\"sk-toggleable__content\"><pre>PassNameTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>ticket</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"b37ce864-e4e3-4708-946c-46dd1a8cb7d9\" type=\"checkbox\" ><label for=\"b37ce864-e4e3-4708-946c-46dd1a8cb7d9\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">TicketTransformer</label><div class=\"sk-toggleable__content\"><pre>TicketTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>cabin</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"4485375b-a383-4b80-91ca-dc9a10465496\" type=\"checkbox\" ><label for=\"4485375b-a383-4b80-91ca-dc9a10465496\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CabinTransformer</label><div class=\"sk-toggleable__content\"><pre>CabinTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>others</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"f859cd86-c626-4814-8504-f8f7f3864c81\" type=\"checkbox\" ><label for=\"f859cd86-c626-4814-8504-f8f7f3864c81\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ColumnSelectTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnSelectTransformer(columns=[&#x27;pclass&#x27;, &#x27;sex&#x27;, &#x27;age&#x27;, &#x27;sibsp&#x27;, &#x27;parch&#x27;,\n",
       "                                 &#x27;fare&#x27;, &#x27;embarked&#x27;])</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"495c2451-2b87-44b9-a70b-bb3c2f4858dc\" type=\"checkbox\" ><label for=\"495c2451-2b87-44b9-a70b-bb3c2f4858dc\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">LowerTransformer</label><div class=\"sk-toggleable__content\"><pre>LowerTransformer()</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('raw_data_processor',\n",
       "                 featureUnion(transformer_list=[('pass_name',\n",
       "                                                 PassNameTransformer()),\n",
       "                                                ('ticket', TicketTransformer()),\n",
       "                                                ('cabin', CabinTransformer()),\n",
       "                                                ('others',\n",
       "                                                 ColumnSelectTransformer(columns=['pclass',\n",
       "                                                                                  'sex',\n",
       "                                                                                  'age',\n",
       "                                                                                  'sibsp',\n",
       "                                                                                  'parch',\n",
       "                                                                                  'fare',\n",
       "                                                                                  'embarked']))])),\n",
       "                ('lower_text_values', LowerTransformer())])"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "from copy import deepcopy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "impute_plan2 = deepcopy(impute_plan)\n",
    "wgt = 0\n",
    "impute_plan2['features_plan']['ticket_str']['catboost_kwgs']['class_weights'] = ((df2['ticket_str'].value_counts().max() / df2['ticket_str'].value_counts()) ** wgt).to_dict()\n",
    "impute_plan2['features_plan']['cabin_str']['catboost_kwgs']['class_weights'] = ((df2['cabin_str'].value_counts().max() / df2['cabin_str'].value_counts()) ** wgt).to_dict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "pl3 = Pipeline(steps = [\n",
    "    ('aaa', pl), \n",
    "    ('catboost_imputor', CatBoostImputor(impute_plan2,  ['fare', 'embarked', 'age', 'ticket_num', 'ticket_str', 'cabin_num', 'cabin_str']))\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 {color: black;background-color: white;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 pre{padding: 0;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-toggleable {background-color: white;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-estimator:hover {background-color: #d4ebff;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-item {z-index: 1;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-parallel::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-parallel-item:only-child::after {width: 0;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;position: relative;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-label-container {position: relative;z-index: 2;text-align: center;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-c0842952-f6f5-4d6e-a8b1-386fcabcefd8\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;aaa&#x27;,\n",
       "                 Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                                  featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                                  PassNameTransformer()),\n",
       "                                                                 (&#x27;ticket&#x27;,\n",
       "                                                                  TicketTransformer()),\n",
       "                                                                 (&#x27;cabin&#x27;,\n",
       "                                                                  CabinTransformer()),\n",
       "                                                                 (&#x27;others&#x27;,\n",
       "                                                                  ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                                   &#x27;sex&#x27;,\n",
       "                                                                                                   &#x27;age&#x27;,\n",
       "                                                                                                   &#x27;sibsp&#x27;,\n",
       "                                                                                                   &#x27;parch&#x27;,\n",
       "                                                                                                   &#x27;fare&#x27;,\n",
       "                                                                                                   &#x27;embarked&#x27;]))])),\n",
       "                                 (&#x27;lower_text_values&#x27;, LowerTransformer())])),\n",
       "                (&#x27;catboost_imp...\n",
       "                                                        &#x27;cabin_num&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                                        &#x27;cabin_str&#x27;: &lt;function fill_unk at 0x7f84ce895dc0&gt;,\n",
       "                                                        &#x27;embarked&#x27;: &lt;function fill_popular at 0x7f84cdd9a670&gt;,\n",
       "                                                        &#x27;fare&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                                        &#x27;ticket_num&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                                        &#x27;ticket_str&#x27;: &lt;function fill_unk at 0x7f84ce895dc0&gt;}},\n",
       "                                 predictors_order=[&#x27;fare&#x27;, &#x27;embarked&#x27;, &#x27;age&#x27;,\n",
       "                                                   &#x27;ticket_num&#x27;, &#x27;ticket_str&#x27;,\n",
       "                                                   &#x27;cabin_num&#x27;,\n",
       "                                                   &#x27;cabin_str&#x27;]))])</pre><b>Please rerun this cell to show the HTML repr or trust the notebook.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"9842196a-1e67-4fc0-8e2f-706932aa08a2\" type=\"checkbox\" ><label for=\"9842196a-1e67-4fc0-8e2f-706932aa08a2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;aaa&#x27;,\n",
       "                 Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                                  featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                                  PassNameTransformer()),\n",
       "                                                                 (&#x27;ticket&#x27;,\n",
       "                                                                  TicketTransformer()),\n",
       "                                                                 (&#x27;cabin&#x27;,\n",
       "                                                                  CabinTransformer()),\n",
       "                                                                 (&#x27;others&#x27;,\n",
       "                                                                  ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                                   &#x27;sex&#x27;,\n",
       "                                                                                                   &#x27;age&#x27;,\n",
       "                                                                                                   &#x27;sibsp&#x27;,\n",
       "                                                                                                   &#x27;parch&#x27;,\n",
       "                                                                                                   &#x27;fare&#x27;,\n",
       "                                                                                                   &#x27;embarked&#x27;]))])),\n",
       "                                 (&#x27;lower_text_values&#x27;, LowerTransformer())])),\n",
       "                (&#x27;catboost_imp...\n",
       "                                                        &#x27;cabin_num&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                                        &#x27;cabin_str&#x27;: &lt;function fill_unk at 0x7f84ce895dc0&gt;,\n",
       "                                                        &#x27;embarked&#x27;: &lt;function fill_popular at 0x7f84cdd9a670&gt;,\n",
       "                                                        &#x27;fare&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                                        &#x27;ticket_num&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                                        &#x27;ticket_str&#x27;: &lt;function fill_unk at 0x7f84ce895dc0&gt;}},\n",
       "                                 predictors_order=[&#x27;fare&#x27;, &#x27;embarked&#x27;, &#x27;age&#x27;,\n",
       "                                                   &#x27;ticket_num&#x27;, &#x27;ticket_str&#x27;,\n",
       "                                                   &#x27;cabin_num&#x27;,\n",
       "                                                   &#x27;cabin_str&#x27;]))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"7660add5-98c1-43dd-a3fc-01e61660c9fd\" type=\"checkbox\" ><label for=\"7660add5-98c1-43dd-a3fc-01e61660c9fd\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">aaa: Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                 featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                 PassNameTransformer()),\n",
       "                                                (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                                                (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                                                (&#x27;others&#x27;,\n",
       "                                                 ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                  &#x27;sex&#x27;,\n",
       "                                                                                  &#x27;age&#x27;,\n",
       "                                                                                  &#x27;sibsp&#x27;,\n",
       "                                                                                  &#x27;parch&#x27;,\n",
       "                                                                                  &#x27;fare&#x27;,\n",
       "                                                                                  &#x27;embarked&#x27;]))])),\n",
       "                (&#x27;lower_text_values&#x27;, LowerTransformer())])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"655d9d3f-dafb-47ab-ab0d-205ae7493d56\" type=\"checkbox\" ><label for=\"655d9d3f-dafb-47ab-ab0d-205ae7493d56\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">raw_data_processor: featureUnion</label><div class=\"sk-toggleable__content\"><pre>featureUnion(transformer_list=[(&#x27;pass_name&#x27;, PassNameTransformer()),\n",
       "                               (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                               (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                               (&#x27;others&#x27;,\n",
       "                                ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                 &#x27;sex&#x27;, &#x27;age&#x27;,\n",
       "                                                                 &#x27;sibsp&#x27;,\n",
       "                                                                 &#x27;parch&#x27;,\n",
       "                                                                 &#x27;fare&#x27;,\n",
       "                                                                 &#x27;embarked&#x27;]))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>pass_name</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"4310cfd8-0056-4e09-b9b8-5d67cd8d8a1f\" type=\"checkbox\" ><label for=\"4310cfd8-0056-4e09-b9b8-5d67cd8d8a1f\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">PassNameTransformer</label><div class=\"sk-toggleable__content\"><pre>PassNameTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>ticket</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"621ac00f-88df-4d40-bb54-ec97ca15714e\" type=\"checkbox\" ><label for=\"621ac00f-88df-4d40-bb54-ec97ca15714e\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">TicketTransformer</label><div class=\"sk-toggleable__content\"><pre>TicketTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>cabin</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"c62d1ddf-f524-4e8e-8a5f-44e98a77055f\" type=\"checkbox\" ><label for=\"c62d1ddf-f524-4e8e-8a5f-44e98a77055f\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CabinTransformer</label><div class=\"sk-toggleable__content\"><pre>CabinTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>others</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"a9c9cc83-79ee-4b50-b862-3dd5f97c6346\" type=\"checkbox\" ><label for=\"a9c9cc83-79ee-4b50-b862-3dd5f97c6346\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ColumnSelectTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnSelectTransformer(columns=[&#x27;pclass&#x27;, &#x27;sex&#x27;, &#x27;age&#x27;, &#x27;sibsp&#x27;, &#x27;parch&#x27;,\n",
       "                                 &#x27;fare&#x27;, &#x27;embarked&#x27;])</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"041e4990-d593-4958-96a8-9865f6a64d99\" type=\"checkbox\" ><label for=\"041e4990-d593-4958-96a8-9865f6a64d99\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">LowerTransformer</label><div class=\"sk-toggleable__content\"><pre>LowerTransformer()</pre></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"f259b35c-ba71-4088-8bf7-7182780a726e\" type=\"checkbox\" ><label for=\"f259b35c-ba71-4088-8bf7-7182780a726e\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CatBoostImputor</label><div class=\"sk-toggleable__content\"><pre>CatBoostImputor(impute_plan={&#x27;cat_features&#x27;: [&#x27;surname&#x27;, &#x27;forename&#x27;,\n",
       "                                              &#x27;ticket_str&#x27;, &#x27;cabin_str&#x27;, &#x27;sex&#x27;,\n",
       "                                              &#x27;embarked&#x27;],\n",
       "                             &#x27;eval_frac&#x27;: 0.1,\n",
       "                             &#x27;features_plan&#x27;: {&#x27;age&#x27;: {&#x27;catboost_kwgs&#x27;: {&#x27;iterations&#x27;: 200},\n",
       "                                                       &#x27;model&#x27;: &lt;class &#x27;catboost.core.CatBoostRegressor&#x27;&gt;,\n",
       "                                                       &#x27;predictors&#x27;: [&#x27;surname&#x27;,\n",
       "                                                                      &#x27;forename&#x27;,\n",
       "                                                                      &#x27;ticket_str&#x27;,\n",
       "                                                                      &#x27;ticket_num&#x27;,\n",
       "                                                                      &#x27;cabin_str&#x27;,\n",
       "                                                                      &#x27;cabin_num&#x27;,\n",
       "                                                                      &#x27;pclass&#x27;,\n",
       "                                                                      &#x27;sex&#x27;,\n",
       "                                                                      &#x27;sibsp&#x27;,\n",
       "                                                                      &#x27;parch&#x27;,\n",
       "                                                                      &#x27;...\n",
       "                                       &#x27;cabin_num&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                       &#x27;cabin_str&#x27;: &lt;function fill_unk at 0x7f84ce895dc0&gt;,\n",
       "                                       &#x27;embarked&#x27;: &lt;function fill_popular at 0x7f84cdd9a670&gt;,\n",
       "                                       &#x27;fare&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                       &#x27;ticket_num&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                       &#x27;ticket_str&#x27;: &lt;function fill_unk at 0x7f84ce895dc0&gt;}},\n",
       "                predictors_order=[&#x27;fare&#x27;, &#x27;embarked&#x27;, &#x27;age&#x27;, &#x27;ticket_num&#x27;,\n",
       "                                  &#x27;ticket_str&#x27;, &#x27;cabin_num&#x27;, &#x27;cabin_str&#x27;])</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('aaa',\n",
       "                 Pipeline(steps=[('raw_data_processor',\n",
       "                                  featureUnion(transformer_list=[('pass_name',\n",
       "                                                                  PassNameTransformer()),\n",
       "                                                                 ('ticket',\n",
       "                                                                  TicketTransformer()),\n",
       "                                                                 ('cabin',\n",
       "                                                                  CabinTransformer()),\n",
       "                                                                 ('others',\n",
       "                                                                  ColumnSelectTransformer(columns=['pclass',\n",
       "                                                                                                   'sex',\n",
       "                                                                                                   'age',\n",
       "                                                                                                   'sibsp',\n",
       "                                                                                                   'parch',\n",
       "                                                                                                   'fare',\n",
       "                                                                                                   'embarked']))])),\n",
       "                                 ('lower_text_values', LowerTransformer())])),\n",
       "                ('catboost_imp...\n",
       "                                                        'cabin_num': <function fill_mean at 0x7f84cdd9a550>,\n",
       "                                                        'cabin_str': <function fill_unk at 0x7f84ce895dc0>,\n",
       "                                                        'embarked': <function fill_popular at 0x7f84cdd9a670>,\n",
       "                                                        'fare': <function fill_mean at 0x7f84cdd9a550>,\n",
       "                                                        'ticket_num': <function fill_mean at 0x7f84cdd9a550>,\n",
       "                                                        'ticket_str': <function fill_unk at 0x7f84ce895dc0>}},\n",
       "                                 predictors_order=['fare', 'embarked', 'age',\n",
       "                                                   'ticket_num', 'ticket_str',\n",
       "                                                   'cabin_num',\n",
       "                                                   'cabin_str']))])"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "pl2 = Pipeline(steps=[\n",
    "    ('raw_data_processor', col_processor),\n",
    "    ('lower_text_values', LowerTransformer()),\n",
    "    ('catboost_imputor', CatBoostImputor(impute_plan2,  ['fare', 'embarked', 'age', 'ticket_num', 'ticket_str', 'cabin_num', 'cabin_str']))\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 {color: black;background-color: white;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 pre{padding: 0;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-toggleable {background-color: white;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-estimator:hover {background-color: #d4ebff;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-item {z-index: 1;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-parallel::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-parallel-item:only-child::after {width: 0;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;position: relative;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-label-container {position: relative;z-index: 2;text-align: center;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-9bacd87f-22f8-46c9-81a5-5f5cbb81fa48\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                 featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                 PassNameTransformer()),\n",
       "                                                (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                                                (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                                                (&#x27;others&#x27;,\n",
       "                                                 ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                  &#x27;sex&#x27;,\n",
       "                                                                                  &#x27;age&#x27;,\n",
       "                                                                                  &#x27;sibsp&#x27;,\n",
       "                                                                                  &#x27;parch&#x27;,\n",
       "                                                                                  &#x27;fare&#x27;,\n",
       "                                                                                  &#x27;embarked&#x27;]))])),\n",
       "                (&#x27;lower_text_values&#x27;, LowerTransformer()),\n",
       "                (&#x27;catboost_imputor&#x27;,\n",
       "                 CatBoostImputor(impu...\n",
       "                                                        &#x27;cabin_num&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                                        &#x27;cabin_str&#x27;: &lt;function fill_unk at 0x7f84ce895dc0&gt;,\n",
       "                                                        &#x27;embarked&#x27;: &lt;function fill_popular at 0x7f84cdd9a670&gt;,\n",
       "                                                        &#x27;fare&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                                        &#x27;ticket_num&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                                        &#x27;ticket_str&#x27;: &lt;function fill_unk at 0x7f84ce895dc0&gt;}},\n",
       "                                 predictors_order=[&#x27;fare&#x27;, &#x27;embarked&#x27;, &#x27;age&#x27;,\n",
       "                                                   &#x27;ticket_num&#x27;, &#x27;ticket_str&#x27;,\n",
       "                                                   &#x27;cabin_num&#x27;,\n",
       "                                                   &#x27;cabin_str&#x27;]))])</pre><b>Please rerun this cell to show the HTML repr or trust the notebook.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"73073e5b-ee63-462f-8662-42c1878e4aaa\" type=\"checkbox\" ><label for=\"73073e5b-ee63-462f-8662-42c1878e4aaa\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;raw_data_processor&#x27;,\n",
       "                 featureUnion(transformer_list=[(&#x27;pass_name&#x27;,\n",
       "                                                 PassNameTransformer()),\n",
       "                                                (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                                                (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                                                (&#x27;others&#x27;,\n",
       "                                                 ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                                  &#x27;sex&#x27;,\n",
       "                                                                                  &#x27;age&#x27;,\n",
       "                                                                                  &#x27;sibsp&#x27;,\n",
       "                                                                                  &#x27;parch&#x27;,\n",
       "                                                                                  &#x27;fare&#x27;,\n",
       "                                                                                  &#x27;embarked&#x27;]))])),\n",
       "                (&#x27;lower_text_values&#x27;, LowerTransformer()),\n",
       "                (&#x27;catboost_imputor&#x27;,\n",
       "                 CatBoostImputor(impu...\n",
       "                                                        &#x27;cabin_num&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                                        &#x27;cabin_str&#x27;: &lt;function fill_unk at 0x7f84ce895dc0&gt;,\n",
       "                                                        &#x27;embarked&#x27;: &lt;function fill_popular at 0x7f84cdd9a670&gt;,\n",
       "                                                        &#x27;fare&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                                        &#x27;ticket_num&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                                        &#x27;ticket_str&#x27;: &lt;function fill_unk at 0x7f84ce895dc0&gt;}},\n",
       "                                 predictors_order=[&#x27;fare&#x27;, &#x27;embarked&#x27;, &#x27;age&#x27;,\n",
       "                                                   &#x27;ticket_num&#x27;, &#x27;ticket_str&#x27;,\n",
       "                                                   &#x27;cabin_num&#x27;,\n",
       "                                                   &#x27;cabin_str&#x27;]))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"508c3e9f-cd46-478b-957e-a98f242cc4f2\" type=\"checkbox\" ><label for=\"508c3e9f-cd46-478b-957e-a98f242cc4f2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">raw_data_processor: featureUnion</label><div class=\"sk-toggleable__content\"><pre>featureUnion(transformer_list=[(&#x27;pass_name&#x27;, PassNameTransformer()),\n",
       "                               (&#x27;ticket&#x27;, TicketTransformer()),\n",
       "                               (&#x27;cabin&#x27;, CabinTransformer()),\n",
       "                               (&#x27;others&#x27;,\n",
       "                                ColumnSelectTransformer(columns=[&#x27;pclass&#x27;,\n",
       "                                                                 &#x27;sex&#x27;, &#x27;age&#x27;,\n",
       "                                                                 &#x27;sibsp&#x27;,\n",
       "                                                                 &#x27;parch&#x27;,\n",
       "                                                                 &#x27;fare&#x27;,\n",
       "                                                                 &#x27;embarked&#x27;]))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>pass_name</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"7369212d-dcd7-44a5-903e-88736458657f\" type=\"checkbox\" ><label for=\"7369212d-dcd7-44a5-903e-88736458657f\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">PassNameTransformer</label><div class=\"sk-toggleable__content\"><pre>PassNameTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>ticket</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"c8693148-1475-4666-836d-1338e1f47d48\" type=\"checkbox\" ><label for=\"c8693148-1475-4666-836d-1338e1f47d48\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">TicketTransformer</label><div class=\"sk-toggleable__content\"><pre>TicketTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>cabin</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"91e136e2-fe2e-4bef-9259-4ce9ccd154c0\" type=\"checkbox\" ><label for=\"91e136e2-fe2e-4bef-9259-4ce9ccd154c0\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CabinTransformer</label><div class=\"sk-toggleable__content\"><pre>CabinTransformer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>others</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"e596c966-4b3c-4382-ab97-dbaf5c5628cf\" type=\"checkbox\" ><label for=\"e596c966-4b3c-4382-ab97-dbaf5c5628cf\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ColumnSelectTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnSelectTransformer(columns=[&#x27;pclass&#x27;, &#x27;sex&#x27;, &#x27;age&#x27;, &#x27;sibsp&#x27;, &#x27;parch&#x27;,\n",
       "                                 &#x27;fare&#x27;, &#x27;embarked&#x27;])</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"3167449d-b637-47ad-8bd9-66f2b2838bef\" type=\"checkbox\" ><label for=\"3167449d-b637-47ad-8bd9-66f2b2838bef\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">LowerTransformer</label><div class=\"sk-toggleable__content\"><pre>LowerTransformer()</pre></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"3b1a05a5-a78e-4c23-99c5-151e8e03b90d\" type=\"checkbox\" ><label for=\"3b1a05a5-a78e-4c23-99c5-151e8e03b90d\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CatBoostImputor</label><div class=\"sk-toggleable__content\"><pre>CatBoostImputor(impute_plan={&#x27;cat_features&#x27;: [&#x27;surname&#x27;, &#x27;forename&#x27;,\n",
       "                                              &#x27;ticket_str&#x27;, &#x27;cabin_str&#x27;, &#x27;sex&#x27;,\n",
       "                                              &#x27;embarked&#x27;],\n",
       "                             &#x27;eval_frac&#x27;: 0.1,\n",
       "                             &#x27;features_plan&#x27;: {&#x27;age&#x27;: {&#x27;catboost_kwgs&#x27;: {&#x27;iterations&#x27;: 200},\n",
       "                                                       &#x27;model&#x27;: &lt;class &#x27;catboost.core.CatBoostRegressor&#x27;&gt;,\n",
       "                                                       &#x27;predictors&#x27;: [&#x27;surname&#x27;,\n",
       "                                                                      &#x27;forename&#x27;,\n",
       "                                                                      &#x27;ticket_str&#x27;,\n",
       "                                                                      &#x27;ticket_num&#x27;,\n",
       "                                                                      &#x27;cabin_str&#x27;,\n",
       "                                                                      &#x27;cabin_num&#x27;,\n",
       "                                                                      &#x27;pclass&#x27;,\n",
       "                                                                      &#x27;sex&#x27;,\n",
       "                                                                      &#x27;sibsp&#x27;,\n",
       "                                                                      &#x27;parch&#x27;,\n",
       "                                                                      &#x27;...\n",
       "                                       &#x27;cabin_num&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                       &#x27;cabin_str&#x27;: &lt;function fill_unk at 0x7f84ce895dc0&gt;,\n",
       "                                       &#x27;embarked&#x27;: &lt;function fill_popular at 0x7f84cdd9a670&gt;,\n",
       "                                       &#x27;fare&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                       &#x27;ticket_num&#x27;: &lt;function fill_mean at 0x7f84cdd9a550&gt;,\n",
       "                                       &#x27;ticket_str&#x27;: &lt;function fill_unk at 0x7f84ce895dc0&gt;}},\n",
       "                predictors_order=[&#x27;fare&#x27;, &#x27;embarked&#x27;, &#x27;age&#x27;, &#x27;ticket_num&#x27;,\n",
       "                                  &#x27;ticket_str&#x27;, &#x27;cabin_num&#x27;, &#x27;cabin_str&#x27;])</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('raw_data_processor',\n",
       "                 featureUnion(transformer_list=[('pass_name',\n",
       "                                                 PassNameTransformer()),\n",
       "                                                ('ticket', TicketTransformer()),\n",
       "                                                ('cabin', CabinTransformer()),\n",
       "                                                ('others',\n",
       "                                                 ColumnSelectTransformer(columns=['pclass',\n",
       "                                                                                  'sex',\n",
       "                                                                                  'age',\n",
       "                                                                                  'sibsp',\n",
       "                                                                                  'parch',\n",
       "                                                                                  'fare',\n",
       "                                                                                  'embarked']))])),\n",
       "                ('lower_text_values', LowerTransformer()),\n",
       "                ('catboost_imputor',\n",
       "                 CatBoostImputor(impu...\n",
       "                                                        'cabin_num': <function fill_mean at 0x7f84cdd9a550>,\n",
       "                                                        'cabin_str': <function fill_unk at 0x7f84ce895dc0>,\n",
       "                                                        'embarked': <function fill_popular at 0x7f84cdd9a670>,\n",
       "                                                        'fare': <function fill_mean at 0x7f84cdd9a550>,\n",
       "                                                        'ticket_num': <function fill_mean at 0x7f84cdd9a550>,\n",
       "                                                        'ticket_str': <function fill_unk at 0x7f84ce895dc0>}},\n",
       "                                 predictors_order=['fare', 'embarked', 'age',\n",
       "                                                   'ticket_num', 'ticket_str',\n",
       "                                                   'cabin_num',\n",
       "                                                   'cabin_str']))])"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====Converting columns types====\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "de4734bc1d284483af58bebaaa94eb85",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prepare data to fit, feature: fare\n",
      "Fit model: fare\n",
      "Learning rate set to 0.275418\n",
      "0:\tlearn: 64.7716052\ttest: 66.1606066\tbest: 66.1606066 (0)\ttotal: 46.9ms\tremaining: 9.33s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 58.0748847\ttest: 60.2768871\tbest: 60.1887846 (15)\ttotal: 1.56s\tremaining: 4.56s\n",
      "Stopped by overfitting detector  (50 iterations wait)\n",
      "\n",
      "bestTest = 60.1887846\n",
      "bestIteration = 15\n",
      "\n",
      "Shrink model to first 16 iterations.\n",
      "Predicted for 134 unknown examples\n",
      "Imputed done for fare\n",
      "Prepare data to fit, feature: embarked\n",
      "Fit model: embarked\n",
      "Learning rate set to 0.218473\n",
      "0:\tlearn: 0.9044038\ttest: 0.9048636\tbest: 0.9048636 (0)\ttotal: 60ms\tremaining: 11.9s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 0.5333882\ttest: 0.5412056\tbest: 0.5412056 (50)\ttotal: 3.15s\tremaining: 9.2s\n",
      "100:\tlearn: 0.5228426\ttest: 0.5369545\tbest: 0.5369545 (100)\ttotal: 6.08s\tremaining: 5.96s\n",
      "150:\tlearn: 0.5167845\ttest: 0.5355787\tbest: 0.5354669 (140)\ttotal: 9.01s\tremaining: 2.92s\n",
      "199:\tlearn: 0.5111682\ttest: 0.5352499\tbest: 0.5351492 (191)\ttotal: 12s\tremaining: 0us\n",
      "\n",
      "bestTest = 0.5351491726\n",
      "bestIteration = 191\n",
      "\n",
      "Shrink model to first 192 iterations.\n",
      "Predicted for 250 unknown examples\n",
      "Imputed done for embarked\n",
      "Prepare data to fit, feature: age\n",
      "Fit model: age\n",
      "Learning rate set to 0.274032\n",
      "0:\tlearn: 17.5600497\ttest: 17.5481511\tbest: 17.5481511 (0)\ttotal: 27.6ms\tremaining: 5.5s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 16.3259788\ttest: 16.3955363\tbest: 16.3949363 (40)\ttotal: 1.13s\tremaining: 3.3s\n",
      "100:\tlearn: 16.2444509\ttest: 16.3918999\tbest: 16.3883138 (73)\ttotal: 2.33s\tremaining: 2.29s\n",
      "150:\tlearn: 16.1731610\ttest: 16.3988955\tbest: 16.3877958 (126)\ttotal: 3.38s\tremaining: 1.1s\n",
      "Stopped by overfitting detector  (50 iterations wait)\n",
      "\n",
      "bestTest = 16.38779579\n",
      "bestIteration = 126\n",
      "\n",
      "Shrink model to first 127 iterations.\n",
      "Predicted for 3292 unknown examples\n",
      "Imputed done for age\n",
      "Prepare data to fit, feature: ticket_num\n",
      "Fit model: ticket_num\n",
      "Learning rate set to 0.260893\n",
      "0:\tlearn: 614895.7356933\ttest: 627042.1296637\tbest: 627042.1296637 (0)\ttotal: 11.6ms\tremaining: 2.31s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 71274.9155187\ttest: 75429.4915013\tbest: 73597.0573244 (41)\ttotal: 890ms\tremaining: 2.6s\n",
      "Stopped by overfitting detector  (50 iterations wait)\n",
      "\n",
      "bestTest = 73597.05732\n",
      "bestIteration = 41\n",
      "\n",
      "Shrink model to first 42 iterations.\n",
      "Predicted for 29278 unknown examples\n",
      "Imputed done for ticket_num\n",
      "Prepare data to fit, feature: ticket_str\n",
      "Fit model: ticket_str\n",
      "Learning rate set to 0.21245\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:\tlearn: 1.3090017\ttest: 1.3783086\tbest: 1.3783086 (0)\ttotal: 2.93s\tremaining: 9m 42s\n",
      "50:\tlearn: 0.0524073\ttest: 0.0332282\tbest: 0.0332282 (50)\ttotal: 3m 5s\tremaining: 9m 2s\n",
      "100:\tlearn: 0.0335230\ttest: 0.0220470\tbest: 0.0220470 (100)\ttotal: 6m 12s\tremaining: 6m 5s\n",
      "150:\tlearn: 0.0243197\ttest: 0.0179098\tbest: 0.0179098 (150)\ttotal: 9m 57s\tremaining: 3m 13s\n",
      "199:\tlearn: 0.0192411\ttest: 0.0163181\tbest: 0.0163181 (199)\ttotal: 13m 34s\tremaining: 0us\n",
      "\n",
      "bestTest = 0.01631812971\n",
      "bestIteration = 199\n",
      "\n",
      "Predicted for 75345 unknown examples\n",
      "Imputed done for ticket_str\n",
      "Prepare data to fit, feature: cabin_num\n",
      "Fit model: cabin_num\n",
      "Learning rate set to 0.230503\n",
      "0:\tlearn: 4498.9583529\ttest: 4541.2841963\tbest: 4541.2841963 (0)\ttotal: 57.5ms\tremaining: 11.4s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 117.8343451\ttest: 116.6287392\tbest: 116.6287392 (50)\ttotal: 683ms\tremaining: 1.99s\n",
      "100:\tlearn: 69.9594958\ttest: 67.1111864\tbest: 67.1111864 (100)\ttotal: 1.37s\tremaining: 1.34s\n",
      "150:\tlearn: 51.1203784\ttest: 48.3114028\tbest: 48.3114028 (150)\ttotal: 1.91s\tremaining: 621ms\n",
      "199:\tlearn: 42.9797373\ttest: 40.3670093\tbest: 40.3667782 (198)\ttotal: 2.44s\tremaining: 0us\n",
      "\n",
      "bestTest = 40.3667782\n",
      "bestIteration = 198\n",
      "\n",
      "Shrink model to first 199 iterations.\n",
      "Predicted for 67866 unknown examples\n",
      "Imputed done for cabin_num\n",
      "Prepare data to fit, feature: cabin_str\n",
      "Fit model: cabin_str\n",
      "Learning rate set to 0.213579\n",
      "0:\tlearn: 0.8504391\ttest: 0.8464051\tbest: 0.8464051 (0)\ttotal: 98.1ms\tremaining: 19.5s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning: Overfitting detector is active, thus evaluation metric is calculated on every iteration. 'metric_period' is ignored for evaluation metric.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50:\tlearn: 0.0019538\ttest: 0.0015575\tbest: 0.0015575 (50)\ttotal: 5.43s\tremaining: 15.9s\n",
      "100:\tlearn: 0.0007435\ttest: 0.0005515\tbest: 0.0005515 (100)\ttotal: 11.4s\tremaining: 11.2s\n",
      "150:\tlearn: 0.0004178\ttest: 0.0002958\tbest: 0.0002958 (150)\ttotal: 18s\tremaining: 5.83s\n",
      "199:\tlearn: 0.0002864\ttest: 0.0001953\tbest: 0.0001953 (199)\ttotal: 24.3s\tremaining: 0us\n",
      "\n",
      "bestTest = 0.000195257891\n",
      "bestIteration = 199\n",
      "\n",
      "Predicted for 67866 unknown examples\n",
      "Imputed done for cabin_str\n",
      "Imputed done for all dataset, num null left: surname       0\n",
      "forename      0\n",
      "ticket_str    0\n",
      "ticket_num    0\n",
      "cabin_str     0\n",
      "cabin_num     0\n",
      "pclass        0\n",
      "sex           0\n",
      "age           0\n",
      "sibsp         0\n",
      "parch         0\n",
      "fare          0\n",
      "embarked      0\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "clean_X_train2 = pl2.fit_transform(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Pipeline'"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl2.__class__.__name__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [],
   "source": [
    "a = CatBoostImputor(impute_plan, predictors_order=['fare', 'embarked', 'age', 'ticket_num', 'ticket_str', 'cabin_num', 'cabin_str'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "from time import time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1642568848.073945"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "time()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "from time import time, sleep\n",
    "class PipelineLogger:\n",
    "    def __init__(self):\n",
    "        pass\n",
    "        \n",
    "    def log_start(self):\n",
    "        self.start_time = time()\n",
    "        print(f'======== {self.__class__.__name__} - START ========')\n",
    "        return None\n",
    "        \n",
    "    def log_finish(self):\n",
    "        self.duration = time() - self.start_time\n",
    "        print(f'======== {self.__class__.__name__} - FINISH =======> Take: {self.duration:.6f}(s)')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "class JustSleep(PipelineLogger):\n",
    "    def __init__(self):\n",
    "        pass\n",
    "    \n",
    "    def sleep(self):\n",
    "        self.log_start()\n",
    "        sleep(1)\n",
    "        self.log_finish()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "js = JustSleep()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "======== JustSleep - START ========\n",
      "======== JustSleep - FINISH =======> Take: 1.005406(s)\n"
     ]
    }
   ],
   "source": [
    "js.sleep()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Submit"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model based on catboost imputed data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test['Survived'] = model_clean.predict(X_test_clean)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>pclass</th>\n",
       "      <th>name</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>sibsp</th>\n",
       "      <th>parch</th>\n",
       "      <th>ticket</th>\n",
       "      <th>fare</th>\n",
       "      <th>cabin</th>\n",
       "      <th>embarked</th>\n",
       "      <th>Survived</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>passengerid</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>100000</th>\n",
       "      <td>3</td>\n",
       "      <td>Holliday, Daniel</td>\n",
       "      <td>male</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>24745</td>\n",
       "      <td>63.01</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100001</th>\n",
       "      <td>3</td>\n",
       "      <td>Nguyen, Lorraine</td>\n",
       "      <td>female</td>\n",
       "      <td>53.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13264</td>\n",
       "      <td>5.81</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100002</th>\n",
       "      <td>1</td>\n",
       "      <td>Harris, Heather</td>\n",
       "      <td>female</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>25990</td>\n",
       "      <td>38.91</td>\n",
       "      <td>B15315</td>\n",
       "      <td>C</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100003</th>\n",
       "      <td>2</td>\n",
       "      <td>Larsen, Eric</td>\n",
       "      <td>male</td>\n",
       "      <td>25.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>314011</td>\n",
       "      <td>12.93</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100004</th>\n",
       "      <td>1</td>\n",
       "      <td>Cleary, Sarah</td>\n",
       "      <td>female</td>\n",
       "      <td>17.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>26203</td>\n",
       "      <td>26.89</td>\n",
       "      <td>B22515</td>\n",
       "      <td>C</td>\n",
       "      <td>1</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199995</th>\n",
       "      <td>3</td>\n",
       "      <td>Cash, Cheryle</td>\n",
       "      <td>female</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7686</td>\n",
       "      <td>10.12</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199996</th>\n",
       "      <td>1</td>\n",
       "      <td>Brown, Howard</td>\n",
       "      <td>male</td>\n",
       "      <td>59.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>13004</td>\n",
       "      <td>68.31</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199997</th>\n",
       "      <td>3</td>\n",
       "      <td>Lightfoot, Cameron</td>\n",
       "      <td>male</td>\n",
       "      <td>47.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4383317</td>\n",
       "      <td>10.87</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199998</th>\n",
       "      <td>1</td>\n",
       "      <td>Jacobsen, Margaret</td>\n",
       "      <td>female</td>\n",
       "      <td>49.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>PC 26988</td>\n",
       "      <td>29.68</td>\n",
       "      <td>B20828</td>\n",
       "      <td>C</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199999</th>\n",
       "      <td>1</td>\n",
       "      <td>Fishback, Joanna</td>\n",
       "      <td>female</td>\n",
       "      <td>41.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>PC 41824</td>\n",
       "      <td>195.41</td>\n",
       "      <td>E13345</td>\n",
       "      <td>C</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100000 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             pclass                name     sex   age  sibsp  parch    ticket  \\\n",
       "passengerid                                                                     \n",
       "100000            3    Holliday, Daniel    male  19.0      0      0     24745   \n",
       "100001            3    Nguyen, Lorraine  female  53.0      0      0     13264   \n",
       "100002            1     Harris, Heather  female  19.0      0      0     25990   \n",
       "100003            2        Larsen, Eric    male  25.0      0      0    314011   \n",
       "100004            1       Cleary, Sarah  female  17.0      0      2     26203   \n",
       "...             ...                 ...     ...   ...    ...    ...       ...   \n",
       "199995            3       Cash, Cheryle  female  27.0      0      0      7686   \n",
       "199996            1       Brown, Howard    male  59.0      1      0     13004   \n",
       "199997            3  Lightfoot, Cameron    male  47.0      0      0   4383317   \n",
       "199998            1  Jacobsen, Margaret  female  49.0      1      2  PC 26988   \n",
       "199999            1    Fishback, Joanna  female  41.0      0      2  PC 41824   \n",
       "\n",
       "               fare   cabin embarked  Survived  \n",
       "passengerid                                     \n",
       "100000        63.01     NaN        S         0  \n",
       "100001         5.81     NaN        S         0  \n",
       "100002        38.91  B15315        C         1  \n",
       "100003        12.93     NaN        S         0  \n",
       "100004        26.89  B22515        C         1  \n",
       "...             ...     ...      ...       ...  \n",
       "199995        10.12     NaN        Q         1  \n",
       "199996        68.31     NaN        S         0  \n",
       "199997        10.87     NaN        S         0  \n",
       "199998        29.68  B20828        C         1  \n",
       "199999       195.41  E13345        C         1  \n",
       "\n",
       "[100000 rows x 11 columns]"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "m = X_test['Survived'].to_frame().reset_index()\n",
    "m.columns = ['PassengerId', 'Survived']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "m.to_csv('submission/v1.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model based on simple imputed data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.8.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}