beginner-titanic-challenge.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "edfab134",
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5",
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:49.492494Z",
     "iopub.status.busy": "2023-05-30T18:20:49.492063Z",
     "iopub.status.idle": "2023-05-30T18:20:49.506499Z",
     "shell.execute_reply": "2023-05-30T18:20:49.505619Z"
    },
    "papermill": {
     "duration": 0.040075,
     "end_time": "2023-05-30T18:20:49.508728",
     "exception": false,
     "start_time": "2023-05-30T18:20:49.468653",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/kaggle/input/titanic/train.csv\n",
      "/kaggle/input/titanic/test.csv\n",
      "/kaggle/input/titanic/gender_submission.csv\n"
     ]
    }
   ],
   "source": [
    "# This Python 3 environment comes with many helpful analytics libraries installed\n",
    "# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n",
    "# For example, here's several helpful packages to load\n",
    "\n",
    "import numpy as np # linear algebra\n",
    "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n",
    "\n",
    "# Input data files are available in the read-only \"../input/\" directory\n",
    "# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n",
    "\n",
    "import os\n",
    "for dirname, _, filenames in os.walk('/kaggle/input'):\n",
    "    for filename in filenames:\n",
    "        print(os.path.join(dirname, filename))\n",
    "\n",
    "# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n",
    "# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a1a8fb1c",
   "metadata": {
    "papermill": {
     "duration": 0.018679,
     "end_time": "2023-05-30T18:20:49.547143",
     "exception": false,
     "start_time": "2023-05-30T18:20:49.528464",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **1. Read in Input and Basic Analysis**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "00413bc5",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:49.587583Z",
     "iopub.status.busy": "2023-05-30T18:20:49.586523Z",
     "iopub.status.idle": "2023-05-30T18:20:49.619193Z",
     "shell.execute_reply": "2023-05-30T18:20:49.617865Z"
    },
    "papermill": {
     "duration": 0.05562,
     "end_time": "2023-05-30T18:20:49.621811",
     "exception": false,
     "start_time": "2023-05-30T18:20:49.566191",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(891, 12)\n"
     ]
    }
   ],
   "source": [
    "# read in the input train data and check the first 5 rows\n",
    "train_data = pd.read_csv(\"/kaggle/input/titanic/train.csv\")\n",
    "print(train_data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "ba1dc5fe",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:49.662457Z",
     "iopub.status.busy": "2023-05-30T18:20:49.661770Z",
     "iopub.status.idle": "2023-05-30T18:20:49.703542Z",
     "shell.execute_reply": "2023-05-30T18:20:49.702613Z"
    },
    "papermill": {
     "duration": 0.064677,
     "end_time": "2023-05-30T18:20:49.705921",
     "exception": false,
     "start_time": "2023-05-30T18:20:49.641244",
     "status": "completed"
    },
    "tags": []
   },
   "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>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Braund, Mr. Owen Harris</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Cumings, Mrs. John Bradley (Florence Briggs Th...</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Heikkinen, Miss. Laina</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Futrelle, Mrs. Jacques Heath (Lily May Peel)</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Allen, Mr. William Henry</td>\n",
       "      <td>male</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>373450</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived  Pclass  \\\n",
       "0            1         0       3   \n",
       "1            2         1       1   \n",
       "2            3         1       3   \n",
       "3            4         1       1   \n",
       "4            5         0       3   \n",
       "\n",
       "                                                Name     Sex   Age  SibSp  \\\n",
       "0                            Braund, Mr. Owen Harris    male  22.0      1   \n",
       "1  Cumings, Mrs. John Bradley (Florence Briggs Th...  female  38.0      1   \n",
       "2                             Heikkinen, Miss. Laina  female  26.0      0   \n",
       "3       Futrelle, Mrs. Jacques Heath (Lily May Peel)  female  35.0      1   \n",
       "4                           Allen, Mr. William Henry    male  35.0      0   \n",
       "\n",
       "   Parch            Ticket     Fare Cabin Embarked  \n",
       "0      0         A/5 21171   7.2500   NaN        S  \n",
       "1      0          PC 17599  71.2833   C85        C  \n",
       "2      0  STON/O2. 3101282   7.9250   NaN        S  \n",
       "3      0            113803  53.1000  C123        S  \n",
       "4      0            373450   8.0500   NaN        S  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "66c106c4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:49.748133Z",
     "iopub.status.busy": "2023-05-30T18:20:49.746787Z",
     "iopub.status.idle": "2023-05-30T18:20:49.768855Z",
     "shell.execute_reply": "2023-05-30T18:20:49.767584Z"
    },
    "papermill": {
     "duration": 0.045715,
     "end_time": "2023-05-30T18:20:49.771343",
     "exception": false,
     "start_time": "2023-05-30T18:20:49.725628",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(418, 11)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# read in the input test data and check the first 5 rows\n",
    "test_data = pd.read_csv(\"/kaggle/input/titanic/test.csv\")\n",
    "test_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "b8418634",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:49.812934Z",
     "iopub.status.busy": "2023-05-30T18:20:49.811776Z",
     "iopub.status.idle": "2023-05-30T18:20:49.829469Z",
     "shell.execute_reply": "2023-05-30T18:20:49.828576Z"
    },
    "papermill": {
     "duration": 0.041012,
     "end_time": "2023-05-30T18:20:49.831823",
     "exception": false,
     "start_time": "2023-05-30T18:20:49.790811",
     "status": "completed"
    },
    "tags": []
   },
   "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>892</td>\n",
       "      <td>3</td>\n",
       "      <td>Kelly, Mr. James</td>\n",
       "      <td>male</td>\n",
       "      <td>34.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>330911</td>\n",
       "      <td>7.8292</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>893</td>\n",
       "      <td>3</td>\n",
       "      <td>Wilkes, Mrs. James (Ellen Needs)</td>\n",
       "      <td>female</td>\n",
       "      <td>47.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>363272</td>\n",
       "      <td>7.0000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>894</td>\n",
       "      <td>2</td>\n",
       "      <td>Myles, Mr. Thomas Francis</td>\n",
       "      <td>male</td>\n",
       "      <td>62.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>240276</td>\n",
       "      <td>9.6875</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>895</td>\n",
       "      <td>3</td>\n",
       "      <td>Wirz, Mr. Albert</td>\n",
       "      <td>male</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>315154</td>\n",
       "      <td>8.6625</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>896</td>\n",
       "      <td>3</td>\n",
       "      <td>Hirvonen, Mrs. Alexander (Helga E Lindqvist)</td>\n",
       "      <td>female</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3101298</td>\n",
       "      <td>12.2875</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Pclass                                          Name     Sex  \\\n",
       "0          892       3                              Kelly, Mr. James    male   \n",
       "1          893       3              Wilkes, Mrs. James (Ellen Needs)  female   \n",
       "2          894       2                     Myles, Mr. Thomas Francis    male   \n",
       "3          895       3                              Wirz, Mr. Albert    male   \n",
       "4          896       3  Hirvonen, Mrs. Alexander (Helga E Lindqvist)  female   \n",
       "\n",
       "    Age  SibSp  Parch   Ticket     Fare Cabin Embarked  \n",
       "0  34.5      0      0   330911   7.8292   NaN        Q  \n",
       "1  47.0      1      0   363272   7.0000   NaN        S  \n",
       "2  62.0      0      0   240276   9.6875   NaN        Q  \n",
       "3  27.0      0      0   315154   8.6625   NaN        S  \n",
       "4  22.0      1      1  3101298  12.2875   NaN        S  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "1d950e12",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:49.873949Z",
     "iopub.status.busy": "2023-05-30T18:20:49.873264Z",
     "iopub.status.idle": "2023-05-30T18:20:49.881027Z",
     "shell.execute_reply": "2023-05-30T18:20:49.879977Z"
    },
    "papermill": {
     "duration": 0.031634,
     "end_time": "2023-05-30T18:20:49.883401",
     "exception": false,
     "start_time": "2023-05-30T18:20:49.851767",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PassengerId      int64\n",
       "Survived         int64\n",
       "Pclass           int64\n",
       "Name            object\n",
       "Sex             object\n",
       "Age            float64\n",
       "SibSp            int64\n",
       "Parch            int64\n",
       "Ticket          object\n",
       "Fare           float64\n",
       "Cabin           object\n",
       "Embarked        object\n",
       "dtype: object"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "320ba7a2",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:49.926124Z",
     "iopub.status.busy": "2023-05-30T18:20:49.925693Z",
     "iopub.status.idle": "2023-05-30T18:20:49.933706Z",
     "shell.execute_reply": "2023-05-30T18:20:49.932911Z"
    },
    "papermill": {
     "duration": 0.032325,
     "end_time": "2023-05-30T18:20:49.935873",
     "exception": false,
     "start_time": "2023-05-30T18:20:49.903548",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PassengerId      int64\n",
       "Pclass           int64\n",
       "Name            object\n",
       "Sex             object\n",
       "Age            float64\n",
       "SibSp            int64\n",
       "Parch            int64\n",
       "Ticket          object\n",
       "Fare           float64\n",
       "Cabin           object\n",
       "Embarked        object\n",
       "dtype: object"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_data.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "a35a8ad7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:49.979031Z",
     "iopub.status.busy": "2023-05-30T18:20:49.978241Z",
     "iopub.status.idle": "2023-05-30T18:20:49.997843Z",
     "shell.execute_reply": "2023-05-30T18:20:49.996346Z"
    },
    "papermill": {
     "duration": 0.043864,
     "end_time": "2023-05-30T18:20:50.000224",
     "exception": false,
     "start_time": "2023-05-30T18:20:49.956360",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PassengerId    891\n",
      "Survived         2\n",
      "Pclass           3\n",
      "Name           891\n",
      "Sex              2\n",
      "Age             88\n",
      "SibSp            7\n",
      "Parch            7\n",
      "Ticket         681\n",
      "Fare           248\n",
      "Cabin          147\n",
      "Embarked         3\n",
      "dtype: int64\n",
      "\n",
      "\n",
      "PassengerId    418\n",
      "Pclass           3\n",
      "Name           418\n",
      "Sex              2\n",
      "Age             79\n",
      "SibSp            7\n",
      "Parch            8\n",
      "Ticket         363\n",
      "Fare           169\n",
      "Cabin           76\n",
      "Embarked         3\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# check if there is any missing values\n",
    "print(train_data.nunique())\n",
    "print('\\n')\n",
    "print(test_data.nunique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "816355c4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:50.043246Z",
     "iopub.status.busy": "2023-05-30T18:20:50.042793Z",
     "iopub.status.idle": "2023-05-30T18:20:50.049768Z",
     "shell.execute_reply": "2023-05-30T18:20:50.048594Z"
    },
    "papermill": {
     "duration": 0.032037,
     "end_time": "2023-05-30T18:20:50.052442",
     "exception": false,
     "start_time": "2023-05-30T18:20:50.020405",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# define a function to get a dictionary of feature/value in test data but missing in train data \n",
    "def get_test_only(train_data, test_data):\n",
    "    test_only_dict = {}\n",
    "    for col in test_data.columns:\n",
    "        if col == 'PassengerId':\n",
    "            continue\n",
    "        train_set = set(train_data[col].unique())\n",
    "        test_set = set(test_data[col].unique())\n",
    "        test_only_list = list(test_set.difference(train_set))\n",
    "        test_only_dict[col] = test_only_list\n",
    "    return test_only_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "ee9fb9bf",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:50.097372Z",
     "iopub.status.busy": "2023-05-30T18:20:50.096512Z",
     "iopub.status.idle": "2023-05-30T18:20:50.108226Z",
     "shell.execute_reply": "2023-05-30T18:20:50.107001Z"
    },
    "papermill": {
     "duration": 0.037363,
     "end_time": "2023-05-30T18:20:50.110890",
     "exception": false,
     "start_time": "2023-05-30T18:20:50.073527",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pclass : []\n",
      "Sex : []\n",
      "Age : [0.17, 0.33, 67.0, 38.5, 11.5, 76.0, nan, 18.5, 22.5, 26.5, 60.5]\n",
      "SibSp : []\n",
      "Parch : [9]\n",
      "Fare : [3.1708, 7.0, 136.7792, 9.6875, 7.8208, 8.9625, 7.2833, 13.9, 13.775, 15.5792, 7.7208, 9.325, 12.1833, 12.7375, 15.0333, 25.7417, 25.7, 27.4458, 28.5375, 31.5, 31.6792, 31.6833, 7.5792, 7.85, 42.5, 45.5, nan, 10.7083, 57.75, 60.0, 75.2417, 82.2667, 7.7792, 7.575]\n",
      "Embarked : []\n"
     ]
    }
   ],
   "source": [
    "# print out unexpected features with distinct values only in test data\n",
    "test_only_dict = get_test_only(train_data, test_data)\n",
    "for key in test_only_dict.keys():\n",
    "    expected_features = ['Name', 'Ticket', 'Cabin']\n",
    "    if key in expected_features:\n",
    "        continue\n",
    "    print(key, \": {}\".format(test_only_dict[key]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "28d5cfa7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:50.154183Z",
     "iopub.status.busy": "2023-05-30T18:20:50.153366Z",
     "iopub.status.idle": "2023-05-30T18:20:50.217154Z",
     "shell.execute_reply": "2023-05-30T18:20:50.215969Z"
    },
    "papermill": {
     "duration": 0.088585,
     "end_time": "2023-05-30T18:20:50.219955",
     "exception": false,
     "start_time": "2023-05-30T18:20:50.131370",
     "status": "completed"
    },
    "tags": []
   },
   "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>count</th>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>891</td>\n",
       "      <td>891</td>\n",
       "      <td>714.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>891</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>204</td>\n",
       "      <td>889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>891</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>681</td>\n",
       "      <td>NaN</td>\n",
       "      <td>147</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Braund, Mr. Owen Harris</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>347082</td>\n",
       "      <td>NaN</td>\n",
       "      <td>B96 B98</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>577</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4</td>\n",
       "      <td>644</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>446.000000</td>\n",
       "      <td>0.383838</td>\n",
       "      <td>2.308642</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>29.699118</td>\n",
       "      <td>0.523008</td>\n",
       "      <td>0.381594</td>\n",
       "      <td>NaN</td>\n",
       "      <td>32.204208</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>257.353842</td>\n",
       "      <td>0.486592</td>\n",
       "      <td>0.836071</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>14.526497</td>\n",
       "      <td>1.102743</td>\n",
       "      <td>0.806057</td>\n",
       "      <td>NaN</td>\n",
       "      <td>49.693429</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.420000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>223.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>20.125000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.910400</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>446.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>14.454200</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>668.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>38.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>891.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>512.329200</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        PassengerId    Survived      Pclass                     Name   Sex  \\\n",
       "count    891.000000  891.000000  891.000000                      891   891   \n",
       "unique          NaN         NaN         NaN                      891     2   \n",
       "top             NaN         NaN         NaN  Braund, Mr. Owen Harris  male   \n",
       "freq            NaN         NaN         NaN                        1   577   \n",
       "mean     446.000000    0.383838    2.308642                      NaN   NaN   \n",
       "std      257.353842    0.486592    0.836071                      NaN   NaN   \n",
       "min        1.000000    0.000000    1.000000                      NaN   NaN   \n",
       "25%      223.500000    0.000000    2.000000                      NaN   NaN   \n",
       "50%      446.000000    0.000000    3.000000                      NaN   NaN   \n",
       "75%      668.500000    1.000000    3.000000                      NaN   NaN   \n",
       "max      891.000000    1.000000    3.000000                      NaN   NaN   \n",
       "\n",
       "               Age       SibSp       Parch  Ticket        Fare    Cabin  \\\n",
       "count   714.000000  891.000000  891.000000     891  891.000000      204   \n",
       "unique         NaN         NaN         NaN     681         NaN      147   \n",
       "top            NaN         NaN         NaN  347082         NaN  B96 B98   \n",
       "freq           NaN         NaN         NaN       7         NaN        4   \n",
       "mean     29.699118    0.523008    0.381594     NaN   32.204208      NaN   \n",
       "std      14.526497    1.102743    0.806057     NaN   49.693429      NaN   \n",
       "min       0.420000    0.000000    0.000000     NaN    0.000000      NaN   \n",
       "25%      20.125000    0.000000    0.000000     NaN    7.910400      NaN   \n",
       "50%      28.000000    0.000000    0.000000     NaN   14.454200      NaN   \n",
       "75%      38.000000    1.000000    0.000000     NaN   31.000000      NaN   \n",
       "max      80.000000    8.000000    6.000000     NaN  512.329200      NaN   \n",
       "\n",
       "       Embarked  \n",
       "count       889  \n",
       "unique        3  \n",
       "top           S  \n",
       "freq        644  \n",
       "mean        NaN  \n",
       "std         NaN  \n",
       "min         NaN  \n",
       "25%         NaN  \n",
       "50%         NaN  \n",
       "75%         NaN  \n",
       "max         NaN  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.describe(include = \"all\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e6fce8a1",
   "metadata": {
    "papermill": {
     "duration": 0.020795,
     "end_time": "2023-05-30T18:20:50.261745",
     "exception": false,
     "start_time": "2023-05-30T18:20:50.240950",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "### **Notes:**\n",
    "* **Need to pay attention to 'Parch', since test data has a value which doesn't exist in train data and out of range of train data.**\n",
    "* **'Parch' is # of parents / children aboard the Titanic.**\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f2e86d90",
   "metadata": {
    "papermill": {
     "duration": 0.020523,
     "end_time": "2023-05-30T18:20:50.303194",
     "exception": false,
     "start_time": "2023-05-30T18:20:50.282671",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **2. Pre-processing**\n",
    "## 2.1. Deal with Null"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "4caaab83",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:50.347991Z",
     "iopub.status.busy": "2023-05-30T18:20:50.347163Z",
     "iopub.status.idle": "2023-05-30T18:20:50.357168Z",
     "shell.execute_reply": "2023-05-30T18:20:50.355784Z"
    },
    "papermill": {
     "duration": 0.035243,
     "end_time": "2023-05-30T18:20:50.359419",
     "exception": false,
     "start_time": "2023-05-30T18:20:50.324176",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PassengerId      0\n",
      "Survived         0\n",
      "Pclass           0\n",
      "Name             0\n",
      "Sex              0\n",
      "Age            177\n",
      "SibSp            0\n",
      "Parch            0\n",
      "Ticket           0\n",
      "Fare             0\n",
      "Cabin          687\n",
      "Embarked         2\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# check if there is any missing values\n",
    "print(train_data.isnull().sum())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "962d1616",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:50.404146Z",
     "iopub.status.busy": "2023-05-30T18:20:50.403153Z",
     "iopub.status.idle": "2023-05-30T18:20:50.412741Z",
     "shell.execute_reply": "2023-05-30T18:20:50.411383Z"
    },
    "papermill": {
     "duration": 0.034744,
     "end_time": "2023-05-30T18:20:50.415459",
     "exception": false,
     "start_time": "2023-05-30T18:20:50.380715",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PassengerId      0\n",
      "Pclass           0\n",
      "Name             0\n",
      "Sex              0\n",
      "Age             86\n",
      "SibSp            0\n",
      "Parch            0\n",
      "Ticket           0\n",
      "Fare             1\n",
      "Cabin          327\n",
      "Embarked         0\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(test_data.isnull().sum())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d028f64b",
   "metadata": {
    "papermill": {
     "duration": 0.020713,
     "end_time": "2023-05-30T18:20:50.457188",
     "exception": false,
     "start_time": "2023-05-30T18:20:50.436475",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "### **Notes:**\n",
    "* **Percentage of missing value for 'Cabin' is very high in both train and test data. Any imputation is likely to work poorly for 'Cabin'. It is better to drop this feature for model building.**\n",
    "* **Percentage of missing value for 'Age' is ~20% for both train and test data. Imputation has a chance to work if model requires no NaN. It is not a good idea to drop 'Age' feature since intuitively it should have a significant impact on our target, survival.**\n",
    "* **For the two missings for 'Embarked' in train data, it is safe to drop those two rows.**\n",
    "* **For the only missing for 'Fare' in test data, we can not drop it. If model can not deal with nan, we can impute a value for it.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "051a184c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:50.501620Z",
     "iopub.status.busy": "2023-05-30T18:20:50.501203Z",
     "iopub.status.idle": "2023-05-30T18:20:50.514444Z",
     "shell.execute_reply": "2023-05-30T18:20:50.513216Z"
    },
    "papermill": {
     "duration": 0.038736,
     "end_time": "2023-05-30T18:20:50.517221",
     "exception": false,
     "start_time": "2023-05-30T18:20:50.478485",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "# of missing of 'Embarked' after dropping:  0\n",
      "(889, 10)\n",
      "(418, 9)\n"
     ]
    }
   ],
   "source": [
    "# train data dropping\n",
    "# from intuition, passengerId should not have big impact on the target\n",
    "train_data_drop = train_data.drop(['PassengerId','Cabin'], axis = 1)\n",
    "# drop the two rows with NaN for 'Embarked'\n",
    "train_data_drop = train_data_drop[train_data_drop['Embarked'].notnull()]\n",
    "train_data_drop.reset_index(inplace = True, drop = True)\n",
    "print(\"# of missing of 'Embarked' after dropping: \", train_data_drop['Embarked'].isnull().sum())\n",
    "print(train_data_drop.shape)\n",
    "\n",
    "# test data dropping\n",
    "test_data_drop = test_data.drop(['PassengerId','Cabin'], axis = 1)\n",
    "print(test_data_drop.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "a0e69f98",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:50.563688Z",
     "iopub.status.busy": "2023-05-30T18:20:50.562548Z",
     "iopub.status.idle": "2023-05-30T18:20:50.570836Z",
     "shell.execute_reply": "2023-05-30T18:20:50.569997Z"
    },
    "papermill": {
     "duration": 0.034334,
     "end_time": "2023-05-30T18:20:50.573420",
     "exception": false,
     "start_time": "2023-05-30T18:20:50.539086",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# impute for 'Age' and 'Fare'\n",
    "\n",
    "def numerical_imputer(train_data, test_data, impute_list):\n",
    "    from sklearn.impute import SimpleImputer as si\n",
    "    imputer = si(strategy = 'median')\n",
    "\n",
    "    train_data_impt = pd.DataFrame(imputer.fit_transform(train_data[impute_list]))\n",
    "    test_data_impt = pd.DataFrame(imputer.fit_transform(test_data[impute_list]))\n",
    "\n",
    "    train_data_impt.columns = train_data[impute_list].columns\n",
    "    test_data_impt.columns = test_data[impute_list].columns\n",
    "\n",
    "    train_data_imp = train_data.copy()\n",
    "    test_data_imp = test_data.copy()\n",
    "\n",
    "    train_data_imp[impute_list] = train_data_impt[impute_list]\n",
    "    test_data_imp[impute_list] = test_data_impt[impute_list]\n",
    "\n",
    "    return train_data_imp, test_data_imp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "080fd706",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:50.618266Z",
     "iopub.status.busy": "2023-05-30T18:20:50.617821Z",
     "iopub.status.idle": "2023-05-30T18:20:52.288125Z",
     "shell.execute_reply": "2023-05-30T18:20:52.286877Z"
    },
    "papermill": {
     "duration": 1.696227,
     "end_time": "2023-05-30T18:20:52.290839",
     "exception": false,
     "start_time": "2023-05-30T18:20:50.594612",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.10/site-packages/scipy/__init__.py:146: UserWarning: A NumPy version >=1.16.5 and <1.23.0 is required for this version of SciPy (detected version 1.23.5\n",
      "  warnings.warn(f\"A NumPy version >={np_minversion} and <{np_maxversion}\"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Survived    0\n",
      "Pclass      0\n",
      "Name        0\n",
      "Sex         0\n",
      "Age         0\n",
      "SibSp       0\n",
      "Parch       0\n",
      "Ticket      0\n",
      "Fare        0\n",
      "Embarked    0\n",
      "dtype: int64\n",
      "\n",
      "\n",
      "Pclass      0\n",
      "Name        0\n",
      "Sex         0\n",
      "Age         0\n",
      "SibSp       0\n",
      "Parch       0\n",
      "Ticket      0\n",
      "Fare        0\n",
      "Embarked    0\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "impute_list = ['Age', 'Fare']\n",
    "\n",
    "train_data_imp, test_data_imp = numerical_imputer(train_data_drop, test_data_drop, impute_list)\n",
    "\n",
    "print(train_data_imp.isnull().sum())\n",
    "print('\\n')\n",
    "print(test_data_imp.isnull().sum())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "eeb63643",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:52.336875Z",
     "iopub.status.busy": "2023-05-30T18:20:52.335692Z",
     "iopub.status.idle": "2023-05-30T18:20:52.360038Z",
     "shell.execute_reply": "2023-05-30T18:20:52.358765Z"
    },
    "papermill": {
     "duration": 0.050304,
     "end_time": "2023-05-30T18:20:52.362796",
     "exception": false,
     "start_time": "2023-05-30T18:20:52.312492",
     "status": "completed"
    },
    "tags": []
   },
   "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>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>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Braund, Mr. Owen Harris</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Cumings, Mrs. John Bradley (Florence Briggs Th...</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Heikkinen, Miss. Laina</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Futrelle, Mrs. Jacques Heath (Lily May Peel)</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Allen, Mr. William Henry</td>\n",
       "      <td>male</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>373450</td>\n",
       "      <td>8.0500</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>884</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>Montvila, Rev. Juozas</td>\n",
       "      <td>male</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>211536</td>\n",
       "      <td>13.0000</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>885</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Graham, Miss. Margaret Edith</td>\n",
       "      <td>female</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>112053</td>\n",
       "      <td>30.0000</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>886</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Johnston, Miss. Catherine Helen \"Carrie\"</td>\n",
       "      <td>female</td>\n",
       "      <td>28.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>W./C. 6607</td>\n",
       "      <td>23.4500</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>887</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Behr, Mr. Karl Howell</td>\n",
       "      <td>male</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>111369</td>\n",
       "      <td>30.0000</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>888</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Dooley, Mr. Patrick</td>\n",
       "      <td>male</td>\n",
       "      <td>32.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>370376</td>\n",
       "      <td>7.7500</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>889 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Survived  Pclass                                               Name  \\\n",
       "0           0       3                            Braund, Mr. Owen Harris   \n",
       "1           1       1  Cumings, Mrs. John Bradley (Florence Briggs Th...   \n",
       "2           1       3                             Heikkinen, Miss. Laina   \n",
       "3           1       1       Futrelle, Mrs. Jacques Heath (Lily May Peel)   \n",
       "4           0       3                           Allen, Mr. William Henry   \n",
       "..        ...     ...                                                ...   \n",
       "884         0       2                              Montvila, Rev. Juozas   \n",
       "885         1       1                       Graham, Miss. Margaret Edith   \n",
       "886         0       3           Johnston, Miss. Catherine Helen \"Carrie\"   \n",
       "887         1       1                              Behr, Mr. Karl Howell   \n",
       "888         0       3                                Dooley, Mr. Patrick   \n",
       "\n",
       "        Sex   Age  SibSp  Parch            Ticket     Fare Embarked  \n",
       "0      male  22.0      1      0         A/5 21171   7.2500        S  \n",
       "1    female  38.0      1      0          PC 17599  71.2833        C  \n",
       "2    female  26.0      0      0  STON/O2. 3101282   7.9250        S  \n",
       "3    female  35.0      1      0            113803  53.1000        S  \n",
       "4      male  35.0      0      0            373450   8.0500        S  \n",
       "..      ...   ...    ...    ...               ...      ...      ...  \n",
       "884    male  27.0      0      0            211536  13.0000        S  \n",
       "885  female  19.0      0      0            112053  30.0000        S  \n",
       "886  female  28.0      1      2        W./C. 6607  23.4500        S  \n",
       "887    male  26.0      0      0            111369  30.0000        C  \n",
       "888    male  32.0      0      0            370376   7.7500        Q  \n",
       "\n",
       "[889 rows x 10 columns]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data_imp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "250ac462",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:52.408824Z",
     "iopub.status.busy": "2023-05-30T18:20:52.408439Z",
     "iopub.status.idle": "2023-05-30T18:20:52.431168Z",
     "shell.execute_reply": "2023-05-30T18:20:52.429934Z"
    },
    "papermill": {
     "duration": 0.048914,
     "end_time": "2023-05-30T18:20:52.433767",
     "exception": false,
     "start_time": "2023-05-30T18:20:52.384853",
     "status": "completed"
    },
    "tags": []
   },
   "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>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</td>\n",
       "      <td>Kelly, Mr. James</td>\n",
       "      <td>male</td>\n",
       "      <td>34.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>330911</td>\n",
       "      <td>7.8292</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3</td>\n",
       "      <td>Wilkes, Mrs. James (Ellen Needs)</td>\n",
       "      <td>female</td>\n",
       "      <td>47.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>363272</td>\n",
       "      <td>7.0000</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>Myles, Mr. Thomas Francis</td>\n",
       "      <td>male</td>\n",
       "      <td>62.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>240276</td>\n",
       "      <td>9.6875</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>Wirz, Mr. Albert</td>\n",
       "      <td>male</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>315154</td>\n",
       "      <td>8.6625</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>Hirvonen, Mrs. Alexander (Helga E Lindqvist)</td>\n",
       "      <td>female</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3101298</td>\n",
       "      <td>12.2875</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>413</th>\n",
       "      <td>3</td>\n",
       "      <td>Spector, Mr. Woolf</td>\n",
       "      <td>male</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>A.5. 3236</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>414</th>\n",
       "      <td>1</td>\n",
       "      <td>Oliva y Ocana, Dona. Fermina</td>\n",
       "      <td>female</td>\n",
       "      <td>39.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17758</td>\n",
       "      <td>108.9000</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>415</th>\n",
       "      <td>3</td>\n",
       "      <td>Saether, Mr. Simon Sivertsen</td>\n",
       "      <td>male</td>\n",
       "      <td>38.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>SOTON/O.Q. 3101262</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>416</th>\n",
       "      <td>3</td>\n",
       "      <td>Ware, Mr. Frederick</td>\n",
       "      <td>male</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>359309</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>417</th>\n",
       "      <td>3</td>\n",
       "      <td>Peter, Master. Michael J</td>\n",
       "      <td>male</td>\n",
       "      <td>27.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2668</td>\n",
       "      <td>22.3583</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>418 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Pclass                                          Name     Sex   Age  \\\n",
       "0         3                              Kelly, Mr. James    male  34.5   \n",
       "1         3              Wilkes, Mrs. James (Ellen Needs)  female  47.0   \n",
       "2         2                     Myles, Mr. Thomas Francis    male  62.0   \n",
       "3         3                              Wirz, Mr. Albert    male  27.0   \n",
       "4         3  Hirvonen, Mrs. Alexander (Helga E Lindqvist)  female  22.0   \n",
       "..      ...                                           ...     ...   ...   \n",
       "413       3                            Spector, Mr. Woolf    male  27.0   \n",
       "414       1                  Oliva y Ocana, Dona. Fermina  female  39.0   \n",
       "415       3                  Saether, Mr. Simon Sivertsen    male  38.5   \n",
       "416       3                           Ware, Mr. Frederick    male  27.0   \n",
       "417       3                      Peter, Master. Michael J    male  27.0   \n",
       "\n",
       "     SibSp  Parch              Ticket      Fare Embarked  \n",
       "0        0      0              330911    7.8292        Q  \n",
       "1        1      0              363272    7.0000        S  \n",
       "2        0      0              240276    9.6875        Q  \n",
       "3        0      0              315154    8.6625        S  \n",
       "4        1      1             3101298   12.2875        S  \n",
       "..     ...    ...                 ...       ...      ...  \n",
       "413      0      0           A.5. 3236    8.0500        S  \n",
       "414      0      0            PC 17758  108.9000        C  \n",
       "415      0      0  SOTON/O.Q. 3101262    7.2500        S  \n",
       "416      0      0              359309    8.0500        S  \n",
       "417      1      1                2668   22.3583        C  \n",
       "\n",
       "[418 rows x 9 columns]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_data_imp"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0879854",
   "metadata": {
    "papermill": {
     "duration": 0.022111,
     "end_time": "2023-05-30T18:20:52.478778",
     "exception": false,
     "start_time": "2023-05-30T18:20:52.456667",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## 2.2. Feature Engineering\n",
    "### 2.2.1 Mutual Information Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "79af9513",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:52.526233Z",
     "iopub.status.busy": "2023-05-30T18:20:52.525798Z",
     "iopub.status.idle": "2023-05-30T18:20:52.543970Z",
     "shell.execute_reply": "2023-05-30T18:20:52.542621Z"
    },
    "papermill": {
     "duration": 0.045031,
     "end_time": "2023-05-30T18:20:52.546359",
     "exception": false,
     "start_time": "2023-05-30T18:20:52.501328",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pclass        int64\n",
      "Name          int64\n",
      "Sex           int64\n",
      "Age         float64\n",
      "SibSp         int64\n",
      "Parch         int64\n",
      "Ticket        int64\n",
      "Fare        float64\n",
      "Embarked      int64\n",
      "dtype: object\n",
      "\n",
      "\n",
      "Pclass       True\n",
      "Name         True\n",
      "Sex          True\n",
      "Age         False\n",
      "SibSp        True\n",
      "Parch        True\n",
      "Ticket       True\n",
      "Fare        False\n",
      "Embarked     True\n",
      "dtype: bool\n"
     ]
    }
   ],
   "source": [
    "x_train = train_data_imp.drop('Survived', axis = 1)\n",
    "y_train = train_data_imp[['Survived']]\n",
    "\n",
    "# label encoding for categorical variables and get discrete_features mark for MIA\n",
    "x_train_datatype = x_train.dtypes\n",
    "x_train_obj_list = x_train_datatype[x_train_datatype == 'object'].index.to_list()\n",
    "\n",
    "for col in x_train_obj_list:\n",
    "    x_train[col], _ = x_train[col].factorize()\n",
    "    \n",
    "print(x_train.dtypes)\n",
    "print('\\n')\n",
    "discrete_features = x_train.dtypes == 'int'\n",
    "print(discrete_features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "02f61be3",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:52.593424Z",
     "iopub.status.busy": "2023-05-30T18:20:52.592989Z",
     "iopub.status.idle": "2023-05-30T18:20:52.614418Z",
     "shell.execute_reply": "2023-05-30T18:20:52.613210Z"
    },
    "papermill": {
     "duration": 0.048058,
     "end_time": "2023-05-30T18:20:52.617003",
     "exception": false,
     "start_time": "2023-05-30T18:20:52.568945",
     "status": "completed"
    },
    "tags": []
   },
   "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>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>8.0500</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>884</th>\n",
       "      <td>2</td>\n",
       "      <td>884</td>\n",
       "      <td>0</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>676</td>\n",
       "      <td>13.0000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>885</th>\n",
       "      <td>1</td>\n",
       "      <td>885</td>\n",
       "      <td>1</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>677</td>\n",
       "      <td>30.0000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>886</th>\n",
       "      <td>3</td>\n",
       "      <td>886</td>\n",
       "      <td>1</td>\n",
       "      <td>28.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>613</td>\n",
       "      <td>23.4500</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>887</th>\n",
       "      <td>1</td>\n",
       "      <td>887</td>\n",
       "      <td>0</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>678</td>\n",
       "      <td>30.0000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>888</th>\n",
       "      <td>3</td>\n",
       "      <td>888</td>\n",
       "      <td>0</td>\n",
       "      <td>32.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>679</td>\n",
       "      <td>7.7500</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>889 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Pclass  Name  Sex   Age  SibSp  Parch  Ticket     Fare  Embarked\n",
       "0         3     0    0  22.0      1      0       0   7.2500         0\n",
       "1         1     1    1  38.0      1      0       1  71.2833         1\n",
       "2         3     2    1  26.0      0      0       2   7.9250         0\n",
       "3         1     3    1  35.0      1      0       3  53.1000         0\n",
       "4         3     4    0  35.0      0      0       4   8.0500         0\n",
       "..      ...   ...  ...   ...    ...    ...     ...      ...       ...\n",
       "884       2   884    0  27.0      0      0     676  13.0000         0\n",
       "885       1   885    1  19.0      0      0     677  30.0000         0\n",
       "886       3   886    1  28.0      1      2     613  23.4500         0\n",
       "887       1   887    0  26.0      0      0     678  30.0000         1\n",
       "888       3   888    0  32.0      0      0     679   7.7500         2\n",
       "\n",
       "[889 rows x 9 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "fdab5fca",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:52.665024Z",
     "iopub.status.busy": "2023-05-30T18:20:52.664575Z",
     "iopub.status.idle": "2023-05-30T18:20:52.706267Z",
     "shell.execute_reply": "2023-05-30T18:20:52.705087Z"
    },
    "papermill": {
     "duration": 0.069098,
     "end_time": "2023-05-30T18:20:52.708884",
     "exception": false,
     "start_time": "2023-05-30T18:20:52.639786",
     "status": "completed"
    },
    "tags": []
   },
   "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>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>889.000000</td>\n",
       "      <td>889.000000</td>\n",
       "      <td>889.000000</td>\n",
       "      <td>889.000000</td>\n",
       "      <td>889.000000</td>\n",
       "      <td>889.000000</td>\n",
       "      <td>889.000000</td>\n",
       "      <td>889.000000</td>\n",
       "      <td>889.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>2.311586</td>\n",
       "      <td>444.000000</td>\n",
       "      <td>0.350956</td>\n",
       "      <td>29.315152</td>\n",
       "      <td>0.524184</td>\n",
       "      <td>0.382452</td>\n",
       "      <td>306.536558</td>\n",
       "      <td>32.096681</td>\n",
       "      <td>0.362205</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.834700</td>\n",
       "      <td>256.776492</td>\n",
       "      <td>0.477538</td>\n",
       "      <td>12.984932</td>\n",
       "      <td>1.103705</td>\n",
       "      <td>0.806761</td>\n",
       "      <td>196.988587</td>\n",
       "      <td>49.697504</td>\n",
       "      <td>0.636157</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.420000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>222.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>22.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>135.000000</td>\n",
       "      <td>7.895800</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>444.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>285.000000</td>\n",
       "      <td>14.454200</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>666.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>474.000000</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>888.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>679.000000</td>\n",
       "      <td>512.329200</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Pclass        Name         Sex         Age       SibSp       Parch  \\\n",
       "count  889.000000  889.000000  889.000000  889.000000  889.000000  889.000000   \n",
       "mean     2.311586  444.000000    0.350956   29.315152    0.524184    0.382452   \n",
       "std      0.834700  256.776492    0.477538   12.984932    1.103705    0.806761   \n",
       "min      1.000000    0.000000    0.000000    0.420000    0.000000    0.000000   \n",
       "25%      2.000000  222.000000    0.000000   22.000000    0.000000    0.000000   \n",
       "50%      3.000000  444.000000    0.000000   28.000000    0.000000    0.000000   \n",
       "75%      3.000000  666.000000    1.000000   35.000000    1.000000    0.000000   \n",
       "max      3.000000  888.000000    1.000000   80.000000    8.000000    6.000000   \n",
       "\n",
       "           Ticket        Fare    Embarked  \n",
       "count  889.000000  889.000000  889.000000  \n",
       "mean   306.536558   32.096681    0.362205  \n",
       "std    196.988587   49.697504    0.636157  \n",
       "min      0.000000    0.000000    0.000000  \n",
       "25%    135.000000    7.895800    0.000000  \n",
       "50%    285.000000   14.454200    0.000000  \n",
       "75%    474.000000   31.000000    1.000000  \n",
       "max    679.000000  512.329200    2.000000  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "e50222f8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:52.757096Z",
     "iopub.status.busy": "2023-05-30T18:20:52.756621Z",
     "iopub.status.idle": "2023-05-30T18:20:52.764634Z",
     "shell.execute_reply": "2023-05-30T18:20:52.763539Z"
    },
    "papermill": {
     "duration": 0.034918,
     "end_time": "2023-05-30T18:20:52.767017",
     "exception": false,
     "start_time": "2023-05-30T18:20:52.732099",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def get_mi_score(x_train, y_train):\n",
    "    \n",
    "    # label encoding for categorical variables and get discrete_features mark for MIA\n",
    "    x_train_datatype = x_train.dtypes\n",
    "    x_train_obj_list = x_train_datatype[x_train_datatype == 'object'].index.to_list()\n",
    "    for col in x_train_obj_list:\n",
    "        x_train[col], _ = x_train[col].factorize()\n",
    "    discrete_features = x_train.dtypes == 'int'\n",
    "\n",
    "    from sklearn.feature_selection import mutual_info_classif\n",
    "    mi_scores = mutual_info_classif(x_train, y_train, discrete_features = discrete_features)\n",
    "    mi_scores = pd.Series(mi_scores, name = 'MI_score', index = x_train.columns)\n",
    "    mi_scores = mi_scores.sort_values(ascending = False)\n",
    "    return mi_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "a088b898",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:52.815276Z",
     "iopub.status.busy": "2023-05-30T18:20:52.814845Z",
     "iopub.status.idle": "2023-05-30T18:20:52.903172Z",
     "shell.execute_reply": "2023-05-30T18:20:52.901990Z"
    },
    "papermill": {
     "duration": 0.116036,
     "end_time": "2023-05-30T18:20:52.906314",
     "exception": false,
     "start_time": "2023-05-30T18:20:52.790278",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.10/site-packages/sklearn/utils/validation.py:1143: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Name        0.665252\n",
       "Ticket      0.571626\n",
       "Sex         0.149725\n",
       "Fare        0.134773\n",
       "Pclass      0.057127\n",
       "SibSp       0.023490\n",
       "Parch       0.016653\n",
       "Embarked    0.014547\n",
       "Age         0.008246\n",
       "Name: MI_score, dtype: float64"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_mi_score(x_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "106618d5",
   "metadata": {
    "papermill": {
     "duration": 0.02299,
     "end_time": "2023-05-30T18:20:52.953528",
     "exception": false,
     "start_time": "2023-05-30T18:20:52.930538",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "### **Notes:**\n",
    "* **'Name' and 'Ticket' are dominating MIA due to their high values.**\n",
    "* **Normalization before MIA should help.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "d8e5e2f8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:53.002802Z",
     "iopub.status.busy": "2023-05-30T18:20:53.002414Z",
     "iopub.status.idle": "2023-05-30T18:20:53.070889Z",
     "shell.execute_reply": "2023-05-30T18:20:53.069695Z"
    },
    "papermill": {
     "duration": 0.096725,
     "end_time": "2023-05-30T18:20:53.073473",
     "exception": false,
     "start_time": "2023-05-30T18:20:52.976748",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.10/site-packages/sklearn/utils/validation.py:1143: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Sex         0.137264\n",
       "Fare        0.132679\n",
       "Ticket      0.097597\n",
       "Pclass      0.069975\n",
       "SibSp       0.038240\n",
       "Age         0.036157\n",
       "Parch       0.019858\n",
       "Name        0.013109\n",
       "Embarked    0.002228\n",
       "Name: MI_score, dtype: float64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# calculate MIA again with normalization\n",
    "x_train_new = (x_train - x_train.mean(axis = 0)) / x_train.std(axis = 0)\n",
    "mi_score = get_mi_score(x_train_new, y_train)\n",
    "mi_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "6465f761",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:53.123708Z",
     "iopub.status.busy": "2023-05-30T18:20:53.123279Z",
     "iopub.status.idle": "2023-05-30T18:20:54.067530Z",
     "shell.execute_reply": "2023-05-30T18:20:54.066220Z"
    },
    "papermill": {
     "duration": 0.972809,
     "end_time": "2023-05-30T18:20:54.070603",
     "exception": false,
     "start_time": "2023-05-30T18:20:53.097794",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Features')"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# make a bar plot to compare MI scores\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "plt.figure(figsize = (9,5))\n",
    "mi_score = pd.DataFrame(mi_score)\n",
    "sns.set_style('darkgrid')\n",
    "sns.barplot(y = mi_score.index, x = 'MI_score', data = mi_score)\n",
    "\n",
    "import numpy as np\n",
    "index_list = np.arange(0, mi_score.shape[0]).tolist()\n",
    "i = 0\n",
    "for row in mi_score.index:\n",
    "    value = mi_score.loc[row,:].MI_score\n",
    "    value = (round(value,4))\n",
    "    plt.annotate(value, xy = (value, i), color = 'blue')\n",
    "    i += 1\n",
    "\n",
    "plt.title('MI Scores of Selected Features')\n",
    "plt.ylabel('Features')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f131c141",
   "metadata": {
    "papermill": {
     "duration": 0.024303,
     "end_time": "2023-05-30T18:20:54.121801",
     "exception": false,
     "start_time": "2023-05-30T18:20:54.097498",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "### 2.2.2 Principal Component Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "6e3baef6",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:54.174219Z",
     "iopub.status.busy": "2023-05-30T18:20:54.173811Z",
     "iopub.status.idle": "2023-05-30T18:20:54.182133Z",
     "shell.execute_reply": "2023-05-30T18:20:54.180679Z"
    },
    "papermill": {
     "duration": 0.037215,
     "end_time": "2023-05-30T18:20:54.184503",
     "exception": false,
     "start_time": "2023-05-30T18:20:54.147288",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def get_pca(x_train_data, standardize = True):\n",
    "    if standardize:\n",
    "        x_train_data = (x_train_data - x_train_data.mean(axis = 0)) / x_train_data.std(axis = 0)\n",
    "    from sklearn.decomposition import PCA\n",
    "    pca = PCA()\n",
    "    x_pca = pca.fit_transform(x_train_data)\n",
    "    comp_names = [f\"PC{i+1}\" for i in range(x_pca.shape[1])]\n",
    "    x_pca = pd.DataFrame(x_pca, columns = comp_names)\n",
    "    loadings = pd.DataFrame(pca.components_.T, columns = comp_names, index = x_train_data.columns)\n",
    "    return pca, x_pca, loadings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "02af35d7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:54.236180Z",
     "iopub.status.busy": "2023-05-30T18:20:54.235055Z",
     "iopub.status.idle": "2023-05-30T18:20:54.258068Z",
     "shell.execute_reply": "2023-05-30T18:20:54.257064Z"
    },
    "papermill": {
     "duration": 0.05189,
     "end_time": "2023-05-30T18:20:54.260723",
     "exception": false,
     "start_time": "2023-05-30T18:20:54.208833",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# let numerical features and binary categorical features for CPA\n",
    "x_train_new_large_MI = x_train_new[['Sex', 'Fare', 'Pclass', 'Embarked', 'Parch', 'SibSp', 'Age']]\n",
    "pca, x_pca, loadings = get_pca(x_train_new_large_MI, False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "b152e23a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:54.311823Z",
     "iopub.status.busy": "2023-05-30T18:20:54.311382Z",
     "iopub.status.idle": "2023-05-30T18:20:54.347178Z",
     "shell.execute_reply": "2023-05-30T18:20:54.345967Z"
    },
    "papermill": {
     "duration": 0.064307,
     "end_time": "2023-05-30T18:20:54.349881",
     "exception": false,
     "start_time": "2023-05-30T18:20:54.285574",
     "status": "completed"
    },
    "tags": []
   },
   "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>PC1</th>\n",
       "      <th>PC2</th>\n",
       "      <th>PC3</th>\n",
       "      <th>PC4</th>\n",
       "      <th>PC5</th>\n",
       "      <th>PC6</th>\n",
       "      <th>PC7</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.323105</td>\n",
       "      <td>-0.096868</td>\n",
       "      <td>-1.796224</td>\n",
       "      <td>-0.668278</td>\n",
       "      <td>-0.018230</td>\n",
       "      <td>0.073436</td>\n",
       "      <td>0.716493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.093180</td>\n",
       "      <td>-0.106079</td>\n",
       "      <td>0.233211</td>\n",
       "      <td>-1.069208</td>\n",
       "      <td>1.429874</td>\n",
       "      <td>-1.191886</td>\n",
       "      <td>0.745057</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.904317</td>\n",
       "      <td>1.762957</td>\n",
       "      <td>-2.218830</td>\n",
       "      <td>-0.769484</td>\n",
       "      <td>1.412906</td>\n",
       "      <td>-0.232116</td>\n",
       "      <td>0.351183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-1.287662</td>\n",
       "      <td>-0.333047</td>\n",
       "      <td>-0.253189</td>\n",
       "      <td>0.153444</td>\n",
       "      <td>-0.155458</td>\n",
       "      <td>0.272043</td>\n",
       "      <td>0.335319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.499643</td>\n",
       "      <td>-1.525546</td>\n",
       "      <td>0.358029</td>\n",
       "      <td>-0.982139</td>\n",
       "      <td>0.230299</td>\n",
       "      <td>-0.285875</td>\n",
       "      <td>0.224912</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>413</th>\n",
       "      <td>-1.293822</td>\n",
       "      <td>-0.336234</td>\n",
       "      <td>-0.253160</td>\n",
       "      <td>0.150191</td>\n",
       "      <td>-0.152030</td>\n",
       "      <td>0.270793</td>\n",
       "      <td>0.328371</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>414</th>\n",
       "      <td>1.451165</td>\n",
       "      <td>1.933080</td>\n",
       "      <td>1.730819</td>\n",
       "      <td>-0.120645</td>\n",
       "      <td>-0.094676</td>\n",
       "      <td>-0.218557</td>\n",
       "      <td>-0.044762</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>415</th>\n",
       "      <td>-1.327413</td>\n",
       "      <td>0.160644</td>\n",
       "      <td>-0.375783</td>\n",
       "      <td>0.097836</td>\n",
       "      <td>0.565135</td>\n",
       "      <td>0.175021</td>\n",
       "      <td>0.502663</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>416</th>\n",
       "      <td>-1.293822</td>\n",
       "      <td>-0.336234</td>\n",
       "      <td>-0.253160</td>\n",
       "      <td>0.150191</td>\n",
       "      <td>-0.152030</td>\n",
       "      <td>0.270793</td>\n",
       "      <td>0.328371</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>417</th>\n",
       "      <td>-0.224301</td>\n",
       "      <td>-1.000502</td>\n",
       "      <td>0.855191</td>\n",
       "      <td>1.386388</td>\n",
       "      <td>0.744771</td>\n",
       "      <td>0.393260</td>\n",
       "      <td>-0.064520</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>418 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          PC1       PC2       PC3       PC4       PC5       PC6       PC7\n",
       "0   -1.323105 -0.096868 -1.796224 -0.668278 -0.018230  0.073436  0.716493\n",
       "1   -0.093180 -0.106079  0.233211 -1.069208  1.429874 -1.191886  0.745057\n",
       "2   -0.904317  1.762957 -2.218830 -0.769484  1.412906 -0.232116  0.351183\n",
       "3   -1.287662 -0.333047 -0.253189  0.153444 -0.155458  0.272043  0.335319\n",
       "4    0.499643 -1.525546  0.358029 -0.982139  0.230299 -0.285875  0.224912\n",
       "..        ...       ...       ...       ...       ...       ...       ...\n",
       "413 -1.293822 -0.336234 -0.253160  0.150191 -0.152030  0.270793  0.328371\n",
       "414  1.451165  1.933080  1.730819 -0.120645 -0.094676 -0.218557 -0.044762\n",
       "415 -1.327413  0.160644 -0.375783  0.097836  0.565135  0.175021  0.502663\n",
       "416 -1.293822 -0.336234 -0.253160  0.150191 -0.152030  0.270793  0.328371\n",
       "417 -0.224301 -1.000502  0.855191  1.386388  0.744771  0.393260 -0.064520\n",
       "\n",
       "[418 rows x 7 columns]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_data_pca = test_data_imp.drop(['Name','Ticket'], axis = 1)\n",
    "\n",
    "# label encoding for categorical variables and get discrete_features mark for MIA\n",
    "test_data_pca_datatype = test_data_pca.dtypes\n",
    "x_test_obj_list = test_data_pca_datatype[test_data_pca_datatype == 'object'].index.to_list()\n",
    "\n",
    "for col in x_test_obj_list:\n",
    "    test_data_pca[col], _ = test_data_pca[col].factorize()\n",
    "\n",
    "test_data_pca = (test_data_pca - test_data_pca.mean(axis = 0)) / test_data_pca.std(axis = 0)\n",
    "test_data_pca = test_data_pca[['Sex', 'Fare', 'Pclass', 'Embarked', 'Parch', 'SibSp', 'Age']]\n",
    "\n",
    "test_data_pca\n",
    "test_pca = pca.transform(test_data_pca)\n",
    "comp_names = [f\"PC{i+1}\" for i in range(test_pca.shape[1])]\n",
    "test_pca = pd.DataFrame(test_pca, columns = comp_names)\n",
    "test_pca"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "296cdae1",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:54.401803Z",
     "iopub.status.busy": "2023-05-30T18:20:54.401411Z",
     "iopub.status.idle": "2023-05-30T18:20:54.419429Z",
     "shell.execute_reply": "2023-05-30T18:20:54.418292Z"
    },
    "papermill": {
     "duration": 0.046758,
     "end_time": "2023-05-30T18:20:54.421703",
     "exception": false,
     "start_time": "2023-05-30T18:20:54.374945",
     "status": "completed"
    },
    "tags": []
   },
   "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>PC1</th>\n",
       "      <th>PC2</th>\n",
       "      <th>PC3</th>\n",
       "      <th>PC4</th>\n",
       "      <th>PC5</th>\n",
       "      <th>PC6</th>\n",
       "      <th>PC7</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.939008</td>\n",
       "      <td>-0.892816</td>\n",
       "      <td>-0.655269</td>\n",
       "      <td>0.351353</td>\n",
       "      <td>-0.313804</td>\n",
       "      <td>-0.389671</td>\n",
       "      <td>0.201033</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.521326</td>\n",
       "      <td>1.387515</td>\n",
       "      <td>1.168596</td>\n",
       "      <td>-0.170189</td>\n",
       "      <td>0.098668</td>\n",
       "      <td>-0.986022</td>\n",
       "      <td>-0.430100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.458752</td>\n",
       "      <td>-0.508749</td>\n",
       "      <td>0.395480</td>\n",
       "      <td>-1.565440</td>\n",
       "      <td>-0.219207</td>\n",
       "      <td>-0.393572</td>\n",
       "      <td>0.569611</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.312825</td>\n",
       "      <td>1.074614</td>\n",
       "      <td>-0.134086</td>\n",
       "      <td>-0.983530</td>\n",
       "      <td>-0.272010</td>\n",
       "      <td>-1.127345</td>\n",
       "      <td>-0.461386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.308482</td>\n",
       "      <td>0.045600</td>\n",
       "      <td>-0.618348</td>\n",
       "      <td>-0.038230</td>\n",
       "      <td>0.253809</td>\n",
       "      <td>0.103460</td>\n",
       "      <td>0.497433</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>884</th>\n",
       "      <td>-0.770020</td>\n",
       "      <td>0.391131</td>\n",
       "      <td>-0.665433</td>\n",
       "      <td>0.028049</td>\n",
       "      <td>-0.527335</td>\n",
       "      <td>0.093935</td>\n",
       "      <td>-0.396193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>885</th>\n",
       "      <td>0.736229</td>\n",
       "      <td>0.637137</td>\n",
       "      <td>0.206740</td>\n",
       "      <td>-1.397059</td>\n",
       "      <td>-1.312386</td>\n",
       "      <td>-0.456776</td>\n",
       "      <td>-0.922042</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>886</th>\n",
       "      <td>1.240493</td>\n",
       "      <td>-1.582758</td>\n",
       "      <td>-0.143325</td>\n",
       "      <td>-1.250741</td>\n",
       "      <td>0.942910</td>\n",
       "      <td>0.710766</td>\n",
       "      <td>0.273718</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>887</th>\n",
       "      <td>-0.101075</td>\n",
       "      <td>1.184202</td>\n",
       "      <td>0.548752</td>\n",
       "      <td>0.854694</td>\n",
       "      <td>-0.656705</td>\n",
       "      <td>0.179632</td>\n",
       "      <td>-1.274318</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>888</th>\n",
       "      <td>-1.286369</td>\n",
       "      <td>0.073366</td>\n",
       "      <td>2.082900</td>\n",
       "      <td>1.406465</td>\n",
       "      <td>0.680085</td>\n",
       "      <td>0.374087</td>\n",
       "      <td>-0.047802</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>889 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          PC1       PC2       PC3       PC4       PC5       PC6       PC7\n",
       "0   -0.939008 -0.892816 -0.655269  0.351353 -0.313804 -0.389671  0.201033\n",
       "1    1.521326  1.387515  1.168596 -0.170189  0.098668 -0.986022 -0.430100\n",
       "2   -0.458752 -0.508749  0.395480 -1.565440 -0.219207 -0.393572  0.569611\n",
       "3    1.312825  1.074614 -0.134086 -0.983530 -0.272010 -1.127345 -0.461386\n",
       "4   -1.308482  0.045600 -0.618348 -0.038230  0.253809  0.103460  0.497433\n",
       "..        ...       ...       ...       ...       ...       ...       ...\n",
       "884 -0.770020  0.391131 -0.665433  0.028049 -0.527335  0.093935 -0.396193\n",
       "885  0.736229  0.637137  0.206740 -1.397059 -1.312386 -0.456776 -0.922042\n",
       "886  1.240493 -1.582758 -0.143325 -1.250741  0.942910  0.710766  0.273718\n",
       "887 -0.101075  1.184202  0.548752  0.854694 -0.656705  0.179632 -1.274318\n",
       "888 -1.286369  0.073366  2.082900  1.406465  0.680085  0.374087 -0.047802\n",
       "\n",
       "[889 rows x 7 columns]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_pca"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "29ebf94d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:54.474497Z",
     "iopub.status.busy": "2023-05-30T18:20:54.474103Z",
     "iopub.status.idle": "2023-05-30T18:20:54.483151Z",
     "shell.execute_reply": "2023-05-30T18:20:54.481829Z"
    },
    "papermill": {
     "duration": 0.038678,
     "end_time": "2023-05-30T18:20:54.485957",
     "exception": false,
     "start_time": "2023-05-30T18:20:54.447279",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def plot_variance(pca, width=8, dpi=100):\n",
    "    # Create figure\n",
    "    fig, axs = plt.subplots(1, 2)\n",
    "    n = pca.n_components_\n",
    "    grid = np.arange(1, n + 1)\n",
    "    # Explained variance\n",
    "    evr = pca.explained_variance_ratio_\n",
    "    axs[0].bar(grid, evr)\n",
    "    axs[0].set(xlabel=\"Component\", title=\"% Explained Variance\", ylim=(0.0, 1.0))\n",
    "    # Cumulative Variance\n",
    "    cv = np.cumsum(evr)\n",
    "    axs[1].plot(np.r_[0, grid], np.r_[0, cv], \"o-\")\n",
    "    axs[1].set(xlabel=\"Component\", title=\"% Cumulative Variance\", ylim=(0.0, 1.0)    )\n",
    "    # Set up figure\n",
    "    fig.set(figwidth=8, dpi=100)\n",
    "    return axs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "a7ac2f24",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:54.540450Z",
     "iopub.status.busy": "2023-05-30T18:20:54.540070Z",
     "iopub.status.idle": "2023-05-30T18:20:55.154233Z",
     "shell.execute_reply": "2023-05-30T18:20:55.153133Z"
    },
    "papermill": {
     "duration": 0.64454,
     "end_time": "2023-05-30T18:20:55.156840",
     "exception": false,
     "start_time": "2023-05-30T18:20:54.512300",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([<AxesSubplot: title={'center': '% Explained Variance'}, xlabel='Component'>,\n",
       "       <AxesSubplot: title={'center': '% Cumulative Variance'}, xlabel='Component'>],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_variance(pca)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "6b9c2135",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:55.217278Z",
     "iopub.status.busy": "2023-05-30T18:20:55.216834Z",
     "iopub.status.idle": "2023-05-30T18:20:55.234444Z",
     "shell.execute_reply": "2023-05-30T18:20:55.233098Z"
    },
    "papermill": {
     "duration": 0.048735,
     "end_time": "2023-05-30T18:20:55.237184",
     "exception": false,
     "start_time": "2023-05-30T18:20:55.188449",
     "status": "completed"
    },
    "tags": []
   },
   "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>PC1</th>\n",
       "      <th>PC2</th>\n",
       "      <th>PC3</th>\n",
       "      <th>PC4</th>\n",
       "      <th>PC5</th>\n",
       "      <th>PC6</th>\n",
       "      <th>PC7</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Sex</th>\n",
       "      <td>0.397113</td>\n",
       "      <td>-0.081176</td>\n",
       "      <td>0.439289</td>\n",
       "      <td>-0.746534</td>\n",
       "      <td>0.034324</td>\n",
       "      <td>-0.271635</td>\n",
       "      <td>0.102273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fare</th>\n",
       "      <td>0.561736</td>\n",
       "      <td>0.290607</td>\n",
       "      <td>-0.002666</td>\n",
       "      <td>0.296587</td>\n",
       "      <td>-0.312552</td>\n",
       "      <td>0.113987</td>\n",
       "      <td>0.633520</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Pclass</th>\n",
       "      <td>-0.388241</td>\n",
       "      <td>-0.548890</td>\n",
       "      <td>0.108762</td>\n",
       "      <td>-0.003336</td>\n",
       "      <td>0.221167</td>\n",
       "      <td>0.070906</td>\n",
       "      <td>0.694410</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Embarked</th>\n",
       "      <td>0.006039</td>\n",
       "      <td>0.050064</td>\n",
       "      <td>0.849247</td>\n",
       "      <td>0.456191</td>\n",
       "      <td>0.192845</td>\n",
       "      <td>0.078657</td>\n",
       "      <td>-0.157325</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Parch</th>\n",
       "      <td>0.475120</td>\n",
       "      <td>-0.351117</td>\n",
       "      <td>-0.138956</td>\n",
       "      <td>-0.044743</td>\n",
       "      <td>0.372293</td>\n",
       "      <td>0.677743</td>\n",
       "      <td>-0.178129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SibSp</th>\n",
       "      <td>0.386590</td>\n",
       "      <td>-0.418975</td>\n",
       "      <td>-0.190524</td>\n",
       "      <td>0.376532</td>\n",
       "      <td>0.237913</td>\n",
       "      <td>-0.657158</td>\n",
       "      <td>-0.092056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Age</th>\n",
       "      <td>-0.028218</td>\n",
       "      <td>0.553488</td>\n",
       "      <td>-0.135501</td>\n",
       "      <td>-0.053142</td>\n",
       "      <td>0.787289</td>\n",
       "      <td>-0.103994</td>\n",
       "      <td>0.202561</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               PC1       PC2       PC3       PC4       PC5       PC6       PC7\n",
       "Sex       0.397113 -0.081176  0.439289 -0.746534  0.034324 -0.271635  0.102273\n",
       "Fare      0.561736  0.290607 -0.002666  0.296587 -0.312552  0.113987  0.633520\n",
       "Pclass   -0.388241 -0.548890  0.108762 -0.003336  0.221167  0.070906  0.694410\n",
       "Embarked  0.006039  0.050064  0.849247  0.456191  0.192845  0.078657 -0.157325\n",
       "Parch     0.475120 -0.351117 -0.138956 -0.044743  0.372293  0.677743 -0.178129\n",
       "SibSp     0.386590 -0.418975 -0.190524  0.376532  0.237913 -0.657158 -0.092056\n",
       "Age      -0.028218  0.553488 -0.135501 -0.053142  0.787289 -0.103994  0.202561"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loadings"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fcf2ef7f",
   "metadata": {
    "papermill": {
     "duration": 0.025701,
     "end_time": "2023-05-30T18:20:55.289256",
     "exception": false,
     "start_time": "2023-05-30T18:20:55.263555",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "### **Notes:**\n",
    "* **From PCA, we can try including the PCs as input features.**\n",
    "* **From loading info, product of 'Fare' and 'Pclass' can be usefull to represent the ticket status of each person.**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cc8d1e67",
   "metadata": {
    "papermill": {
     "duration": 0.032992,
     "end_time": "2023-05-30T18:20:55.351463",
     "exception": false,
     "start_time": "2023-05-30T18:20:55.318471",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "### 2.2.3 Features Creation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "695af570",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:55.408574Z",
     "iopub.status.busy": "2023-05-30T18:20:55.407534Z",
     "iopub.status.idle": "2023-05-30T18:20:55.419430Z",
     "shell.execute_reply": "2023-05-30T18:20:55.417931Z"
    },
    "papermill": {
     "duration": 0.043698,
     "end_time": "2023-05-30T18:20:55.421955",
     "exception": false,
     "start_time": "2023-05-30T18:20:55.378257",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# define a function to tokenize Name column and split Ticket column\n",
    "def pre_process(df):\n",
    "    df = df.copy()\n",
    "    def normalize_name(x):\n",
    "        return \" \".join([v.strip(\",()[].\\\"'\") for v in x.split()])\n",
    "    def ticket_number(x):\n",
    "        return x.split()[-1]\n",
    "    def ticket_item(x):\n",
    "        items = x.split()\n",
    "        if len(items) == 1:\n",
    "            return 'None'\n",
    "        return \"_\".join(items[0:-1])\n",
    "    \n",
    "    df[\"Name\"] = df[\"Name\"].apply(normalize_name)\n",
    "    df[\"Ticket_number\"] = df[\"Ticket\"].apply(ticket_number)\n",
    "    df[\"Ticket_item\"] = df[\"Ticket\"].apply(ticket_item)\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "ba4f82e0",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:55.480857Z",
     "iopub.status.busy": "2023-05-30T18:20:55.480004Z",
     "iopub.status.idle": "2023-05-30T18:20:55.512607Z",
     "shell.execute_reply": "2023-05-30T18:20:55.511408Z"
    },
    "papermill": {
     "duration": 0.066369,
     "end_time": "2023-05-30T18:20:55.515260",
     "exception": false,
     "start_time": "2023-05-30T18:20:55.448891",
     "status": "completed"
    },
    "tags": []
   },
   "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",
       "      <th>Ticket_number</th>\n",
       "      <th>Ticket_item</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Braund Mr Owen Harris</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>21171</td>\n",
       "      <td>A/5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Cumings Mrs John Bradley Florence Briggs Thayer</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "      <td>17599</td>\n",
       "      <td>PC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Heikkinen Miss Laina</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>3101282</td>\n",
       "      <td>STON/O2.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Futrelle Mrs Jacques Heath Lily May Peel</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "      <td>113803</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Allen Mr William Henry</td>\n",
       "      <td>male</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>373450</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>373450</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived  Pclass  \\\n",
       "0            1         0       3   \n",
       "1            2         1       1   \n",
       "2            3         1       3   \n",
       "3            4         1       1   \n",
       "4            5         0       3   \n",
       "\n",
       "                                              Name     Sex   Age  SibSp  \\\n",
       "0                            Braund Mr Owen Harris    male  22.0      1   \n",
       "1  Cumings Mrs John Bradley Florence Briggs Thayer  female  38.0      1   \n",
       "2                             Heikkinen Miss Laina  female  26.0      0   \n",
       "3         Futrelle Mrs Jacques Heath Lily May Peel  female  35.0      1   \n",
       "4                           Allen Mr William Henry    male  35.0      0   \n",
       "\n",
       "   Parch            Ticket     Fare Cabin Embarked Ticket_number Ticket_item  \n",
       "0      0         A/5 21171   7.2500   NaN        S         21171         A/5  \n",
       "1      0          PC 17599  71.2833   C85        C         17599          PC  \n",
       "2      0  STON/O2. 3101282   7.9250   NaN        S       3101282    STON/O2.  \n",
       "3      0            113803  53.1000  C123        S        113803        None  \n",
       "4      0            373450   8.0500   NaN        S        373450        None  "
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pre1_train_data = pre_process(train_data)\n",
    "pre1_test_data = pre_process(test_data)\n",
    "\n",
    "pre1_train_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "8b90bcef",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:55.574207Z",
     "iopub.status.busy": "2023-05-30T18:20:55.573221Z",
     "iopub.status.idle": "2023-05-30T18:20:55.605433Z",
     "shell.execute_reply": "2023-05-30T18:20:55.604504Z"
    },
    "papermill": {
     "duration": 0.064176,
     "end_time": "2023-05-30T18:20:55.608200",
     "exception": false,
     "start_time": "2023-05-30T18:20:55.544024",
     "status": "completed"
    },
    "tags": []
   },
   "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",
       "      <th>Ticket_number</th>\n",
       "      <th>Ticket_item</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Braund Mr Owen Harris</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>21171</td>\n",
       "      <td>A/5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Cumings Mrs John Bradley Florence Briggs Thayer</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "      <td>17599</td>\n",
       "      <td>PC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Heikkinen Miss Laina</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>3101282</td>\n",
       "      <td>STON/O2.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Futrelle Mrs Jacques Heath Lily May Peel</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "      <td>113803</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Allen Mr William Henry</td>\n",
       "      <td>male</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>373450</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>373450</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>884</th>\n",
       "      <td>887</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>Montvila Rev Juozas</td>\n",
       "      <td>male</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>211536</td>\n",
       "      <td>13.0000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>211536</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>885</th>\n",
       "      <td>888</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Graham Miss Margaret Edith</td>\n",
       "      <td>female</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>112053</td>\n",
       "      <td>30.0000</td>\n",
       "      <td>B42</td>\n",
       "      <td>S</td>\n",
       "      <td>112053</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>886</th>\n",
       "      <td>889</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Johnston Miss Catherine Helen Carrie</td>\n",
       "      <td>female</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>W./C. 6607</td>\n",
       "      <td>23.4500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>6607</td>\n",
       "      <td>W./C.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>887</th>\n",
       "      <td>890</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Behr Mr Karl Howell</td>\n",
       "      <td>male</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>111369</td>\n",
       "      <td>30.0000</td>\n",
       "      <td>C148</td>\n",
       "      <td>C</td>\n",
       "      <td>111369</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>888</th>\n",
       "      <td>891</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Dooley Mr Patrick</td>\n",
       "      <td>male</td>\n",
       "      <td>32.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>370376</td>\n",
       "      <td>7.7500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "      <td>370376</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>889 rows × 14 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     PassengerId  Survived  Pclass  \\\n",
       "0              1         0       3   \n",
       "1              2         1       1   \n",
       "2              3         1       3   \n",
       "3              4         1       1   \n",
       "4              5         0       3   \n",
       "..           ...       ...     ...   \n",
       "884          887         0       2   \n",
       "885          888         1       1   \n",
       "886          889         0       3   \n",
       "887          890         1       1   \n",
       "888          891         0       3   \n",
       "\n",
       "                                                Name     Sex   Age  SibSp  \\\n",
       "0                              Braund Mr Owen Harris    male  22.0      1   \n",
       "1    Cumings Mrs John Bradley Florence Briggs Thayer  female  38.0      1   \n",
       "2                               Heikkinen Miss Laina  female  26.0      0   \n",
       "3           Futrelle Mrs Jacques Heath Lily May Peel  female  35.0      1   \n",
       "4                             Allen Mr William Henry    male  35.0      0   \n",
       "..                                               ...     ...   ...    ...   \n",
       "884                              Montvila Rev Juozas    male  27.0      0   \n",
       "885                       Graham Miss Margaret Edith  female  19.0      0   \n",
       "886             Johnston Miss Catherine Helen Carrie  female   NaN      1   \n",
       "887                              Behr Mr Karl Howell    male  26.0      0   \n",
       "888                                Dooley Mr Patrick    male  32.0      0   \n",
       "\n",
       "     Parch            Ticket     Fare Cabin Embarked Ticket_number Ticket_item  \n",
       "0        0         A/5 21171   7.2500   NaN        S         21171         A/5  \n",
       "1        0          PC 17599  71.2833   C85        C         17599          PC  \n",
       "2        0  STON/O2. 3101282   7.9250   NaN        S       3101282    STON/O2.  \n",
       "3        0            113803  53.1000  C123        S        113803        None  \n",
       "4        0            373450   8.0500   NaN        S        373450        None  \n",
       "..     ...               ...      ...   ...      ...           ...         ...  \n",
       "884      0            211536  13.0000   NaN        S        211536        None  \n",
       "885      0            112053  30.0000   B42        S        112053        None  \n",
       "886      2        W./C. 6607  23.4500   NaN        S          6607       W./C.  \n",
       "887      0            111369  30.0000  C148        C        111369        None  \n",
       "888      0            370376   7.7500   NaN        Q        370376        None  \n",
       "\n",
       "[889 rows x 14 columns]"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pre2_train_data = pre1_train_data[pre1_train_data['Embarked'].notnull()]\n",
    "pre2_train_data.reset_index(inplace = True, drop = True)\n",
    "\n",
    "pre2_train_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "606ff157",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:55.667821Z",
     "iopub.status.busy": "2023-05-30T18:20:55.666862Z",
     "iopub.status.idle": "2023-05-30T18:20:55.701238Z",
     "shell.execute_reply": "2023-05-30T18:20:55.700443Z"
    },
    "papermill": {
     "duration": 0.066542,
     "end_time": "2023-05-30T18:20:55.704250",
     "exception": false,
     "start_time": "2023-05-30T18:20:55.637708",
     "status": "completed"
    },
    "tags": []
   },
   "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",
       "      <th>Ticket_number</th>\n",
       "      <th>Ticket_item</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>892</td>\n",
       "      <td>3</td>\n",
       "      <td>Kelly Mr James</td>\n",
       "      <td>male</td>\n",
       "      <td>34.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>330911</td>\n",
       "      <td>7.8292</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "      <td>330911</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>893</td>\n",
       "      <td>3</td>\n",
       "      <td>Wilkes Mrs James Ellen Needs</td>\n",
       "      <td>female</td>\n",
       "      <td>47.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>363272</td>\n",
       "      <td>7.0000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>363272</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>894</td>\n",
       "      <td>2</td>\n",
       "      <td>Myles Mr Thomas Francis</td>\n",
       "      <td>male</td>\n",
       "      <td>62.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>240276</td>\n",
       "      <td>9.6875</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "      <td>240276</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>895</td>\n",
       "      <td>3</td>\n",
       "      <td>Wirz Mr Albert</td>\n",
       "      <td>male</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>315154</td>\n",
       "      <td>8.6625</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>315154</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>896</td>\n",
       "      <td>3</td>\n",
       "      <td>Hirvonen Mrs Alexander Helga E Lindqvist</td>\n",
       "      <td>female</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3101298</td>\n",
       "      <td>12.2875</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>3101298</td>\n",
       "      <td>None</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>413</th>\n",
       "      <td>1305</td>\n",
       "      <td>3</td>\n",
       "      <td>Spector Mr Woolf</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>A.5. 3236</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>3236</td>\n",
       "      <td>A.5.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>414</th>\n",
       "      <td>1306</td>\n",
       "      <td>1</td>\n",
       "      <td>Oliva y Ocana Dona Fermina</td>\n",
       "      <td>female</td>\n",
       "      <td>39.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17758</td>\n",
       "      <td>108.9000</td>\n",
       "      <td>C105</td>\n",
       "      <td>C</td>\n",
       "      <td>17758</td>\n",
       "      <td>PC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>415</th>\n",
       "      <td>1307</td>\n",
       "      <td>3</td>\n",
       "      <td>Saether Mr Simon Sivertsen</td>\n",
       "      <td>male</td>\n",
       "      <td>38.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>SOTON/O.Q. 3101262</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>3101262</td>\n",
       "      <td>SOTON/O.Q.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>416</th>\n",
       "      <td>1308</td>\n",
       "      <td>3</td>\n",
       "      <td>Ware Mr Frederick</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>359309</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>359309</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>417</th>\n",
       "      <td>1309</td>\n",
       "      <td>3</td>\n",
       "      <td>Peter Master Michael J</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2668</td>\n",
       "      <td>22.3583</td>\n",
       "      <td>NaN</td>\n",
       "      <td>C</td>\n",
       "      <td>2668</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>418 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     PassengerId  Pclass                                      Name     Sex  \\\n",
       "0            892       3                            Kelly Mr James    male   \n",
       "1            893       3              Wilkes Mrs James Ellen Needs  female   \n",
       "2            894       2                   Myles Mr Thomas Francis    male   \n",
       "3            895       3                            Wirz Mr Albert    male   \n",
       "4            896       3  Hirvonen Mrs Alexander Helga E Lindqvist  female   \n",
       "..           ...     ...                                       ...     ...   \n",
       "413         1305       3                          Spector Mr Woolf    male   \n",
       "414         1306       1                Oliva y Ocana Dona Fermina  female   \n",
       "415         1307       3                Saether Mr Simon Sivertsen    male   \n",
       "416         1308       3                         Ware Mr Frederick    male   \n",
       "417         1309       3                    Peter Master Michael J    male   \n",
       "\n",
       "      Age  SibSp  Parch              Ticket      Fare Cabin Embarked  \\\n",
       "0    34.5      0      0              330911    7.8292   NaN        Q   \n",
       "1    47.0      1      0              363272    7.0000   NaN        S   \n",
       "2    62.0      0      0              240276    9.6875   NaN        Q   \n",
       "3    27.0      0      0              315154    8.6625   NaN        S   \n",
       "4    22.0      1      1             3101298   12.2875   NaN        S   \n",
       "..    ...    ...    ...                 ...       ...   ...      ...   \n",
       "413   NaN      0      0           A.5. 3236    8.0500   NaN        S   \n",
       "414  39.0      0      0            PC 17758  108.9000  C105        C   \n",
       "415  38.5      0      0  SOTON/O.Q. 3101262    7.2500   NaN        S   \n",
       "416   NaN      0      0              359309    8.0500   NaN        S   \n",
       "417   NaN      1      1                2668   22.3583   NaN        C   \n",
       "\n",
       "    Ticket_number Ticket_item  \n",
       "0          330911        None  \n",
       "1          363272        None  \n",
       "2          240276        None  \n",
       "3          315154        None  \n",
       "4         3101298        None  \n",
       "..            ...         ...  \n",
       "413          3236        A.5.  \n",
       "414         17758          PC  \n",
       "415       3101262  SOTON/O.Q.  \n",
       "416        359309        None  \n",
       "417          2668        None  \n",
       "\n",
       "[418 rows x 13 columns]"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pre2_train_data = pre2_train_data.join(x_pca)\n",
    "pre2_test_data = pre1_test_data.join(test_pca)\n",
    "pre1_test_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "4ccb31c9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:55.765046Z",
     "iopub.status.busy": "2023-05-30T18:20:55.764090Z",
     "iopub.status.idle": "2023-05-30T18:20:55.799721Z",
     "shell.execute_reply": "2023-05-30T18:20:55.798939Z"
    },
    "papermill": {
     "duration": 0.067955,
     "end_time": "2023-05-30T18:20:55.802679",
     "exception": false,
     "start_time": "2023-05-30T18:20:55.734724",
     "status": "completed"
    },
    "tags": []
   },
   "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",
       "      <th>Ticket_number</th>\n",
       "      <th>Ticket_item</th>\n",
       "      <th>PC1</th>\n",
       "      <th>PC2</th>\n",
       "      <th>PC3</th>\n",
       "      <th>PC4</th>\n",
       "      <th>PC5</th>\n",
       "      <th>PC6</th>\n",
       "      <th>PC7</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>892</td>\n",
       "      <td>3</td>\n",
       "      <td>Kelly Mr James</td>\n",
       "      <td>male</td>\n",
       "      <td>34.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>330911</td>\n",
       "      <td>7.8292</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "      <td>330911</td>\n",
       "      <td>None</td>\n",
       "      <td>-1.323105</td>\n",
       "      <td>-0.096868</td>\n",
       "      <td>-1.796224</td>\n",
       "      <td>-0.668278</td>\n",
       "      <td>-0.018230</td>\n",
       "      <td>0.073436</td>\n",
       "      <td>0.716493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>893</td>\n",
       "      <td>3</td>\n",
       "      <td>Wilkes Mrs James Ellen Needs</td>\n",
       "      <td>female</td>\n",
       "      <td>47.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>363272</td>\n",
       "      <td>7.0000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>363272</td>\n",
       "      <td>None</td>\n",
       "      <td>-0.093180</td>\n",
       "      <td>-0.106079</td>\n",
       "      <td>0.233211</td>\n",
       "      <td>-1.069208</td>\n",
       "      <td>1.429874</td>\n",
       "      <td>-1.191886</td>\n",
       "      <td>0.745057</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>894</td>\n",
       "      <td>2</td>\n",
       "      <td>Myles Mr Thomas Francis</td>\n",
       "      <td>male</td>\n",
       "      <td>62.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>240276</td>\n",
       "      <td>9.6875</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "      <td>240276</td>\n",
       "      <td>None</td>\n",
       "      <td>-0.904317</td>\n",
       "      <td>1.762957</td>\n",
       "      <td>-2.218830</td>\n",
       "      <td>-0.769484</td>\n",
       "      <td>1.412906</td>\n",
       "      <td>-0.232116</td>\n",
       "      <td>0.351183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>895</td>\n",
       "      <td>3</td>\n",
       "      <td>Wirz Mr Albert</td>\n",
       "      <td>male</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>315154</td>\n",
       "      <td>8.6625</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>315154</td>\n",
       "      <td>None</td>\n",
       "      <td>-1.287662</td>\n",
       "      <td>-0.333047</td>\n",
       "      <td>-0.253189</td>\n",
       "      <td>0.153444</td>\n",
       "      <td>-0.155458</td>\n",
       "      <td>0.272043</td>\n",
       "      <td>0.335319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>896</td>\n",
       "      <td>3</td>\n",
       "      <td>Hirvonen Mrs Alexander Helga E Lindqvist</td>\n",
       "      <td>female</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3101298</td>\n",
       "      <td>12.2875</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>3101298</td>\n",
       "      <td>None</td>\n",
       "      <td>0.499643</td>\n",
       "      <td>-1.525546</td>\n",
       "      <td>0.358029</td>\n",
       "      <td>-0.982139</td>\n",
       "      <td>0.230299</td>\n",
       "      <td>-0.285875</td>\n",
       "      <td>0.224912</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>413</th>\n",
       "      <td>1305</td>\n",
       "      <td>3</td>\n",
       "      <td>Spector Mr Woolf</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>A.5. 3236</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>3236</td>\n",
       "      <td>A.5.</td>\n",
       "      <td>-1.293822</td>\n",
       "      <td>-0.336234</td>\n",
       "      <td>-0.253160</td>\n",
       "      <td>0.150191</td>\n",
       "      <td>-0.152030</td>\n",
       "      <td>0.270793</td>\n",
       "      <td>0.328371</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>414</th>\n",
       "      <td>1306</td>\n",
       "      <td>1</td>\n",
       "      <td>Oliva y Ocana Dona Fermina</td>\n",
       "      <td>female</td>\n",
       "      <td>39.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17758</td>\n",
       "      <td>108.9000</td>\n",
       "      <td>C105</td>\n",
       "      <td>C</td>\n",
       "      <td>17758</td>\n",
       "      <td>PC</td>\n",
       "      <td>1.451165</td>\n",
       "      <td>1.933080</td>\n",
       "      <td>1.730819</td>\n",
       "      <td>-0.120645</td>\n",
       "      <td>-0.094676</td>\n",
       "      <td>-0.218557</td>\n",
       "      <td>-0.044762</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>415</th>\n",
       "      <td>1307</td>\n",
       "      <td>3</td>\n",
       "      <td>Saether Mr Simon Sivertsen</td>\n",
       "      <td>male</td>\n",
       "      <td>38.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>SOTON/O.Q. 3101262</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>3101262</td>\n",
       "      <td>SOTON/O.Q.</td>\n",
       "      <td>-1.327413</td>\n",
       "      <td>0.160644</td>\n",
       "      <td>-0.375783</td>\n",
       "      <td>0.097836</td>\n",
       "      <td>0.565135</td>\n",
       "      <td>0.175021</td>\n",
       "      <td>0.502663</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>416</th>\n",
       "      <td>1308</td>\n",
       "      <td>3</td>\n",
       "      <td>Ware Mr Frederick</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>359309</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>359309</td>\n",
       "      <td>None</td>\n",
       "      <td>-1.293822</td>\n",
       "      <td>-0.336234</td>\n",
       "      <td>-0.253160</td>\n",
       "      <td>0.150191</td>\n",
       "      <td>-0.152030</td>\n",
       "      <td>0.270793</td>\n",
       "      <td>0.328371</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>417</th>\n",
       "      <td>1309</td>\n",
       "      <td>3</td>\n",
       "      <td>Peter Master Michael J</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2668</td>\n",
       "      <td>22.3583</td>\n",
       "      <td>NaN</td>\n",
       "      <td>C</td>\n",
       "      <td>2668</td>\n",
       "      <td>None</td>\n",
       "      <td>-0.224301</td>\n",
       "      <td>-1.000502</td>\n",
       "      <td>0.855191</td>\n",
       "      <td>1.386388</td>\n",
       "      <td>0.744771</td>\n",
       "      <td>0.393260</td>\n",
       "      <td>-0.064520</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>418 rows × 20 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     PassengerId  Pclass                                      Name     Sex  \\\n",
       "0            892       3                            Kelly Mr James    male   \n",
       "1            893       3              Wilkes Mrs James Ellen Needs  female   \n",
       "2            894       2                   Myles Mr Thomas Francis    male   \n",
       "3            895       3                            Wirz Mr Albert    male   \n",
       "4            896       3  Hirvonen Mrs Alexander Helga E Lindqvist  female   \n",
       "..           ...     ...                                       ...     ...   \n",
       "413         1305       3                          Spector Mr Woolf    male   \n",
       "414         1306       1                Oliva y Ocana Dona Fermina  female   \n",
       "415         1307       3                Saether Mr Simon Sivertsen    male   \n",
       "416         1308       3                         Ware Mr Frederick    male   \n",
       "417         1309       3                    Peter Master Michael J    male   \n",
       "\n",
       "      Age  SibSp  Parch              Ticket      Fare Cabin Embarked  \\\n",
       "0    34.5      0      0              330911    7.8292   NaN        Q   \n",
       "1    47.0      1      0              363272    7.0000   NaN        S   \n",
       "2    62.0      0      0              240276    9.6875   NaN        Q   \n",
       "3    27.0      0      0              315154    8.6625   NaN        S   \n",
       "4    22.0      1      1             3101298   12.2875   NaN        S   \n",
       "..    ...    ...    ...                 ...       ...   ...      ...   \n",
       "413   NaN      0      0           A.5. 3236    8.0500   NaN        S   \n",
       "414  39.0      0      0            PC 17758  108.9000  C105        C   \n",
       "415  38.5      0      0  SOTON/O.Q. 3101262    7.2500   NaN        S   \n",
       "416   NaN      0      0              359309    8.0500   NaN        S   \n",
       "417   NaN      1      1                2668   22.3583   NaN        C   \n",
       "\n",
       "    Ticket_number Ticket_item       PC1       PC2       PC3       PC4  \\\n",
       "0          330911        None -1.323105 -0.096868 -1.796224 -0.668278   \n",
       "1          363272        None -0.093180 -0.106079  0.233211 -1.069208   \n",
       "2          240276        None -0.904317  1.762957 -2.218830 -0.769484   \n",
       "3          315154        None -1.287662 -0.333047 -0.253189  0.153444   \n",
       "4         3101298        None  0.499643 -1.525546  0.358029 -0.982139   \n",
       "..            ...         ...       ...       ...       ...       ...   \n",
       "413          3236        A.5. -1.293822 -0.336234 -0.253160  0.150191   \n",
       "414         17758          PC  1.451165  1.933080  1.730819 -0.120645   \n",
       "415       3101262  SOTON/O.Q. -1.327413  0.160644 -0.375783  0.097836   \n",
       "416        359309        None -1.293822 -0.336234 -0.253160  0.150191   \n",
       "417          2668        None -0.224301 -1.000502  0.855191  1.386388   \n",
       "\n",
       "          PC5       PC6       PC7  \n",
       "0   -0.018230  0.073436  0.716493  \n",
       "1    1.429874 -1.191886  0.745057  \n",
       "2    1.412906 -0.232116  0.351183  \n",
       "3   -0.155458  0.272043  0.335319  \n",
       "4    0.230299 -0.285875  0.224912  \n",
       "..        ...       ...       ...  \n",
       "413 -0.152030  0.270793  0.328371  \n",
       "414 -0.094676 -0.218557 -0.044762  \n",
       "415  0.565135  0.175021  0.502663  \n",
       "416 -0.152030  0.270793  0.328371  \n",
       "417  0.744771  0.393260 -0.064520  \n",
       "\n",
       "[418 rows x 20 columns]"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pre2_test_data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "103b52bf",
   "metadata": {
    "papermill": {
     "duration": 0.027735,
     "end_time": "2023-05-30T18:20:55.862747",
     "exception": false,
     "start_time": "2023-05-30T18:20:55.835012",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **3. Modeling**\n",
    "## **3.1. tfdf.keras.RandomForestModel**\n",
    "### tfdf.keras.RandomForestModel can take care of NaN values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "8fcca162",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:55.922073Z",
     "iopub.status.busy": "2023-05-30T18:20:55.921122Z",
     "iopub.status.idle": "2023-05-30T18:20:55.929184Z",
     "shell.execute_reply": "2023-05-30T18:20:55.927941Z"
    },
    "papermill": {
     "duration": 0.040433,
     "end_time": "2023-05-30T18:20:55.931466",
     "exception": false,
     "start_time": "2023-05-30T18:20:55.891033",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input features: ['Pclass', 'Sex', 'Age', 'SibSp', 'Parch', 'Fare', 'Cabin', 'Embarked', 'Ticket_number', 'Ticket_item', 'PC1', 'PC2', 'PC3', 'PC4', 'PC5', 'PC6', 'PC7']\n"
     ]
    }
   ],
   "source": [
    "# create a list of features for modeling \n",
    "features = list(pre2_train_data.columns)\n",
    "\n",
    "features.remove('Ticket')\n",
    "features.remove('PassengerId')\n",
    "features.remove('Survived')\n",
    "features.remove('Name')\n",
    "\n",
    "print(\"Input features: {}\".format(features))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "27fa81e9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:20:55.991325Z",
     "iopub.status.busy": "2023-05-30T18:20:55.990337Z",
     "iopub.status.idle": "2023-05-30T18:21:05.121944Z",
     "shell.execute_reply": "2023-05-30T18:21:05.120665Z"
    },
    "papermill": {
     "duration": 9.164616,
     "end_time": "2023-05-30T18:21:05.124792",
     "exception": false,
     "start_time": "2023-05-30T18:20:55.960176",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# convert pandas dataframe to tensorflow dataset\n",
    "import tensorflow as tf\n",
    "import tensorflow_decision_forests as tfdf\n",
    "\n",
    "#def tokenize_names(df, labels = None):\n",
    "#    df['Name'] = tf.strings.split(df['Name'])\n",
    "#    return df, labels\n",
    "\n",
    "# auto hyperparameter tuning can not handle categorical set ['Name']\n",
    "pre2_train_data = pre2_train_data.drop('Name', axis = 1)\n",
    "pre2_test_data = pre2_test_data.drop('Name', axis = 1)\n",
    "#pre2_test_data.head()\n",
    "\n",
    "train_ds = tfdf.keras.pd_dataframe_to_tf_dataset(pre2_train_data, label = 'Survived')\n",
    "test_ds = tfdf.keras.pd_dataframe_to_tf_dataset(pre2_test_data)\n",
    "\n",
    "#type(train_ds)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "75fa0d09",
   "metadata": {
    "papermill": {
     "duration": 0.027531,
     "end_time": "2023-05-30T18:21:05.180505",
     "exception": false,
     "start_time": "2023-05-30T18:21:05.152974",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## 3.1.1. Train model with default parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "7d756aa4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:21:05.239075Z",
     "iopub.status.busy": "2023-05-30T18:21:05.238298Z",
     "iopub.status.idle": "2023-05-30T18:21:16.561454Z",
     "shell.execute_reply": "2023-05-30T18:21:16.560346Z"
    },
    "papermill": {
     "duration": 11.354824,
     "end_time": "2023-05-30T18:21:16.564048",
     "exception": false,
     "start_time": "2023-05-30T18:21:05.209224",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[INFO 2023-05-30T18:21:13.346318194+00:00 kernel.cc:1214] Loading model from path /tmp/tmpniwp5__5/model/ with prefix 00b6bffb544e4c81\n",
      "[INFO 2023-05-30T18:21:13.353282676+00:00 abstract_model.cc:1311] Engine \"GradientBoostedTreesQuickScorerExtended\" built\n",
      "[INFO 2023-05-30T18:21:13.35341627+00:00 kernel.cc:1046] Use fast generic engine\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING: AutoGraph could not transform <function simple_ml_inference_op_with_handle at 0x7e61ad6375b0> and will run it as-is.\n",
      "Please report this to the TensorFlow team. When filing the bug, set the verbosity to 10 (on Linux, `export AUTOGRAPH_VERBOSITY=10`) and attach the full output.\n",
      "Cause: could not get source code\n",
      "To silence this warning, decorate the function with @tf.autograph.experimental.do_not_convert\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7e61fee175b0>"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# train 2nd model: gradient boosted trees model\n",
    "\n",
    "model = tfdf.keras.GradientBoostedTreesModel(verbose = 0,\n",
    "                                            features = [tfdf.keras.FeatureUsage(name = n) for n in features],\n",
    "                                            exclude_non_specified_features = True,\n",
    "                                            random_seed = 1234)\n",
    "model.fit(train_ds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "efbd875a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:21:16.623434Z",
     "iopub.status.busy": "2023-05-30T18:21:16.622786Z",
     "iopub.status.idle": "2023-05-30T18:21:16.629699Z",
     "shell.execute_reply": "2023-05-30T18:21:16.628926Z"
    },
    "papermill": {
     "duration": 0.039478,
     "end_time": "2023-05-30T18:21:16.632188",
     "exception": false,
     "start_time": "2023-05-30T18:21:16.592710",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.7802197933197021  Loss: 0.9438779950141907\n"
     ]
    }
   ],
   "source": [
    "# evalue the accuracy\n",
    "self_evaluation = model.make_inspector().evaluation()\n",
    "print(f\"Accuracy: {self_evaluation.accuracy}  Loss: {self_evaluation.loss}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "8b0b97fe",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:21:16.691702Z",
     "iopub.status.busy": "2023-05-30T18:21:16.690791Z",
     "iopub.status.idle": "2023-05-30T18:21:17.295594Z",
     "shell.execute_reply": "2023-05-30T18:21:17.294331Z"
    },
    "papermill": {
     "duration": 0.637515,
     "end_time": "2023-05-30T18:21:17.298404",
     "exception": false,
     "start_time": "2023-05-30T18:21:16.660889",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8931383490562439"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# valuate the model\n",
    "model.compile(['accuracy'])\n",
    "model_accuracy = model.evaluate(train_ds, return_dict = True, verbose = 0)['accuracy']\n",
    "model_accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "97f2678a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:21:17.360117Z",
     "iopub.status.busy": "2023-05-30T18:21:17.359710Z",
     "iopub.status.idle": "2023-05-30T18:21:17.368668Z",
     "shell.execute_reply": "2023-05-30T18:21:17.367541Z"
    },
    "papermill": {
     "duration": 0.042838,
     "end_time": "2023-05-30T18:21:17.370922",
     "exception": false,
     "start_time": "2023-05-30T18:21:17.328084",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'adapt_subsample_for_maximum_training_duration': False,\n",
       " 'allow_na_conditions': False,\n",
       " 'apply_link_function': True,\n",
       " 'categorical_algorithm': 'CART',\n",
       " 'categorical_set_split_greedy_sampling': 0.1,\n",
       " 'categorical_set_split_max_num_items': -1,\n",
       " 'categorical_set_split_min_item_frequency': 1,\n",
       " 'compute_permutation_variable_importance': False,\n",
       " 'dart_dropout': 0.01,\n",
       " 'early_stopping': 'LOSS_INCREASE',\n",
       " 'early_stopping_initial_iteration': 10,\n",
       " 'early_stopping_num_trees_look_ahead': 30,\n",
       " 'focal_loss_alpha': 0.5,\n",
       " 'focal_loss_gamma': 2.0,\n",
       " 'forest_extraction': 'MART',\n",
       " 'goss_alpha': 0.2,\n",
       " 'goss_beta': 0.1,\n",
       " 'growing_strategy': 'LOCAL',\n",
       " 'honest': False,\n",
       " 'honest_fixed_separation': False,\n",
       " 'honest_ratio_leaf_examples': 0.5,\n",
       " 'in_split_min_examples_check': True,\n",
       " 'keep_non_leaf_label_distribution': True,\n",
       " 'l1_regularization': 0.0,\n",
       " 'l2_categorical_regularization': 1.0,\n",
       " 'l2_regularization': 0.0,\n",
       " 'lambda_loss': 1.0,\n",
       " 'loss': 'DEFAULT',\n",
       " 'max_depth': 6,\n",
       " 'max_num_nodes': None,\n",
       " 'maximum_model_size_in_memory_in_bytes': -1.0,\n",
       " 'maximum_training_duration_seconds': -1.0,\n",
       " 'min_examples': 5,\n",
       " 'missing_value_policy': 'GLOBAL_IMPUTATION',\n",
       " 'num_candidate_attributes': -1,\n",
       " 'num_candidate_attributes_ratio': -1.0,\n",
       " 'num_trees': 300,\n",
       " 'pure_serving_model': False,\n",
       " 'random_seed': 1234,\n",
       " 'sampling_method': 'RANDOM',\n",
       " 'selective_gradient_boosting_ratio': 0.01,\n",
       " 'shrinkage': 0.1,\n",
       " 'sorting_strategy': 'PRESORT',\n",
       " 'sparse_oblique_normalization': None,\n",
       " 'sparse_oblique_num_projections_exponent': None,\n",
       " 'sparse_oblique_projection_density_factor': None,\n",
       " 'sparse_oblique_weights': None,\n",
       " 'split_axis': 'AXIS_ALIGNED',\n",
       " 'subsample': 1.0,\n",
       " 'uplift_min_examples_in_treatment': 5,\n",
       " 'uplift_split_score': 'KULLBACK_LEIBLER',\n",
       " 'use_hessian_gain': False,\n",
       " 'validation_interval_in_trees': 1,\n",
       " 'validation_ratio': 0.1}"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# list the default hyperparameters of the model\n",
    "model.learner_params"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a0013e9b",
   "metadata": {
    "papermill": {
     "duration": 0.028746,
     "end_time": "2023-05-30T18:21:17.428449",
     "exception": false,
     "start_time": "2023-05-30T18:21:17.399703",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## 3.1.2. Train model with automated hyper-parameter tuning\n",
    "### Automatci defination of the hyper-parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "f35c2f4a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:21:17.488107Z",
     "iopub.status.busy": "2023-05-30T18:21:17.487647Z",
     "iopub.status.idle": "2023-05-30T18:21:17.493186Z",
     "shell.execute_reply": "2023-05-30T18:21:17.491992Z"
    },
    "papermill": {
     "duration": 0.038483,
     "end_time": "2023-05-30T18:21:17.495543",
     "exception": false,
     "start_time": "2023-05-30T18:21:17.457060",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Create a Random Search tuner with 100 trials and automatic hp configuration.\n",
    "tuner = tfdf.tuner.RandomSearch(num_trials = 100, use_predefined_hps = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "817bfeeb",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:21:17.556029Z",
     "iopub.status.busy": "2023-05-30T18:21:17.555620Z",
     "iopub.status.idle": "2023-05-30T18:23:03.850858Z",
     "shell.execute_reply": "2023-05-30T18:23:03.849574Z"
    },
    "papermill": {
     "duration": 106.329013,
     "end_time": "2023-05-30T18:23:03.853744",
     "exception": false,
     "start_time": "2023-05-30T18:21:17.524731",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Use /tmp/tmplf_8vqpz as temporary training directory\n",
      "Reading training dataset...\n",
      "Training tensor examples:\n",
      "Features: {'PassengerId': <tf.Tensor 'data_12:0' shape=(None,) dtype=int64>, 'Pclass': <tf.Tensor 'data_13:0' shape=(None,) dtype=int64>, 'Sex': <tf.Tensor 'data_14:0' shape=(None,) dtype=string>, 'Age': <tf.Tensor 'data:0' shape=(None,) dtype=float64>, 'SibSp': <tf.Tensor 'data_15:0' shape=(None,) dtype=int64>, 'Parch': <tf.Tensor 'data_11:0' shape=(None,) dtype=int64>, 'Ticket': <tf.Tensor 'data_16:0' shape=(None,) dtype=string>, 'Fare': <tf.Tensor 'data_3:0' shape=(None,) dtype=float64>, 'Cabin': <tf.Tensor 'data_1:0' shape=(None,) dtype=string>, 'Embarked': <tf.Tensor 'data_2:0' shape=(None,) dtype=string>, 'Ticket_number': <tf.Tensor 'data_18:0' shape=(None,) dtype=string>, 'Ticket_item': <tf.Tensor 'data_17:0' shape=(None,) dtype=string>, 'PC1': <tf.Tensor 'data_4:0' shape=(None,) dtype=float64>, 'PC2': <tf.Tensor 'data_5:0' shape=(None,) dtype=float64>, 'PC3': <tf.Tensor 'data_6:0' shape=(None,) dtype=float64>, 'PC4': <tf.Tensor 'data_7:0' shape=(None,) dtype=float64>, 'PC5': <tf.Tensor 'data_8:0' shape=(None,) dtype=float64>, 'PC6': <tf.Tensor 'data_9:0' shape=(None,) dtype=float64>, 'PC7': <tf.Tensor 'data_10:0' shape=(None,) dtype=float64>}\n",
      "Label: Tensor(\"data_19:0\", shape=(None,), dtype=int64)\n",
      "Weights: None\n",
      "Normalized tensor features:\n",
      " {'PassengerId': SemanticTensor(semantic=<Semantic.NUMERICAL: 1>, tensor=<tf.Tensor 'Cast:0' shape=(None,) dtype=float32>), 'Pclass': SemanticTensor(semantic=<Semantic.NUMERICAL: 1>, tensor=<tf.Tensor 'Cast_1:0' shape=(None,) dtype=float32>), 'Sex': SemanticTensor(semantic=<Semantic.CATEGORICAL: 2>, tensor=<tf.Tensor 'data_14:0' shape=(None,) dtype=string>), 'Age': SemanticTensor(semantic=<Semantic.NUMERICAL: 1>, tensor=<tf.Tensor 'Cast_2:0' shape=(None,) dtype=float32>), 'SibSp': SemanticTensor(semantic=<Semantic.NUMERICAL: 1>, tensor=<tf.Tensor 'Cast_3:0' shape=(None,) dtype=float32>), 'Parch': SemanticTensor(semantic=<Semantic.NUMERICAL: 1>, tensor=<tf.Tensor 'Cast_4:0' shape=(None,) dtype=float32>), 'Ticket': SemanticTensor(semantic=<Semantic.CATEGORICAL: 2>, tensor=<tf.Tensor 'data_16:0' shape=(None,) dtype=string>), 'Fare': SemanticTensor(semantic=<Semantic.NUMERICAL: 1>, tensor=<tf.Tensor 'Cast_5:0' shape=(None,) dtype=float32>), 'Cabin': SemanticTensor(semantic=<Semantic.CATEGORICAL: 2>, tensor=<tf.Tensor 'data_1:0' shape=(None,) dtype=string>), 'Embarked': SemanticTensor(semantic=<Semantic.CATEGORICAL: 2>, tensor=<tf.Tensor 'data_2:0' shape=(None,) dtype=string>), 'Ticket_number': SemanticTensor(semantic=<Semantic.CATEGORICAL: 2>, tensor=<tf.Tensor 'data_18:0' shape=(None,) dtype=string>), 'Ticket_item': SemanticTensor(semantic=<Semantic.CATEGORICAL: 2>, tensor=<tf.Tensor 'data_17:0' shape=(None,) dtype=string>), 'PC1': SemanticTensor(semantic=<Semantic.NUMERICAL: 1>, tensor=<tf.Tensor 'Cast_6:0' shape=(None,) dtype=float32>), 'PC2': SemanticTensor(semantic=<Semantic.NUMERICAL: 1>, tensor=<tf.Tensor 'Cast_7:0' shape=(None,) dtype=float32>), 'PC3': SemanticTensor(semantic=<Semantic.NUMERICAL: 1>, tensor=<tf.Tensor 'Cast_8:0' shape=(None,) dtype=float32>), 'PC4': SemanticTensor(semantic=<Semantic.NUMERICAL: 1>, tensor=<tf.Tensor 'Cast_9:0' shape=(None,) dtype=float32>), 'PC5': SemanticTensor(semantic=<Semantic.NUMERICAL: 1>, tensor=<tf.Tensor 'Cast_10:0' shape=(None,) dtype=float32>), 'PC6': SemanticTensor(semantic=<Semantic.NUMERICAL: 1>, tensor=<tf.Tensor 'Cast_11:0' shape=(None,) dtype=float32>), 'PC7': SemanticTensor(semantic=<Semantic.NUMERICAL: 1>, tensor=<tf.Tensor 'Cast_12:0' shape=(None,) dtype=float32>)}\n",
      "Training dataset read in 0:00:00.441790. Found 889 examples.\n",
      "Training model...\n",
      "Standard output detected as not visible to the user e.g. running in a notebook. Creating a training log redirection. If training gets stuck, try calling tfdf.keras.set_training_logs_redirection(False).\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[INFO 2023-05-30T18:21:18.035942244+00:00 kernel.cc:756] Start Yggdrasil model training\n",
      "[INFO 2023-05-30T18:21:18.036096662+00:00 kernel.cc:757] Collect training examples\n",
      "[INFO 2023-05-30T18:21:18.036222926+00:00 kernel.cc:388] Number of batches: 1\n",
      "[INFO 2023-05-30T18:21:18.036241574+00:00 kernel.cc:389] Number of examples: 889\n",
      "[INFO 2023-05-30T18:21:18.037113732+00:00 data_spec_inference.cc:303] 146 item(s) have been pruned (i.e. they are considered out of dictionary) for the column Cabin (0 item(s) left) because min_value_count=5 and max_number_of_unique_values=2000\n",
      "[INFO 2023-05-30T18:21:18.037332769+00:00 data_spec_inference.cc:303] 672 item(s) have been pruned (i.e. they are considered out of dictionary) for the column Ticket (8 item(s) left) because min_value_count=5 and max_number_of_unique_values=2000\n",
      "[INFO 2023-05-30T18:21:18.037442972+00:00 data_spec_inference.cc:303] 28 item(s) have been pruned (i.e. they are considered out of dictionary) for the column Ticket_item (16 item(s) left) because min_value_count=5 and max_number_of_unique_values=2000\n",
      "[INFO 2023-05-30T18:21:18.037621606+00:00 data_spec_inference.cc:303] 670 item(s) have been pruned (i.e. they are considered out of dictionary) for the column Ticket_number (8 item(s) left) because min_value_count=5 and max_number_of_unique_values=2000\n",
      "[INFO 2023-05-30T18:21:18.0379385+00:00 kernel.cc:774] Training dataset:\n",
      "Number of records: 889\n",
      "Number of columns: 20\n",
      "\n",
      "Number of columns by type:\n",
      "\tNUMERICAL: 13 (65%)\n",
      "\tCATEGORICAL: 7 (35%)\n",
      "\n",
      "Columns:\n",
      "\n",
      "NUMERICAL: 13 (65%)\n",
      "\t0: \"Age\" NUMERICAL num-nas:177 (19.91%) mean:29.6421 min:0.42 max:80 sd:14.4828\n",
      "\t3: \"Fare\" NUMERICAL mean:32.0967 min:0 max:512.329 sd:49.6695\n",
      "\t4: \"PC1\" NUMERICAL mean:8.33813e-10 min:-1.4299 max:6.16277 sd:1.33969\n",
      "\t5: \"PC2\" NUMERICAL mean:2.92754e-09 min:-3.97044 max:4.38826 sd:1.29533\n",
      "\t6: \"PC3\" NUMERICAL mean:-2.34513e-09 min:-2.27406 max:3.18019 sd:1.04558\n",
      "\t7: \"PC4\" NUMERICAL mean:1.22089e-09 min:-1.72108 max:3.69633 sd:0.896969\n",
      "\t8: \"PC5\" NUMERICAL mean:-9.38132e-10 min:-3.14443 max:3.55076 sd:0.842955\n",
      "\t9: \"PC6\" NUMERICAL mean:6.29088e-10 min:-3.35137 max:4.00473 sd:0.744096\n",
      "\t10: \"PC7\" NUMERICAL mean:-2.82936e-10 min:-1.41101 max:5.22875 sd:0.597851\n",
      "\t11: \"Parch\" NUMERICAL mean:0.382452 min:0 max:6 sd:0.806307\n",
      "\t12: \"PassengerId\" NUMERICAL mean:446 min:1 max:891 sd:256.854\n",
      "\t13: \"Pclass\" NUMERICAL mean:2.31159 min:1 max:3 sd:0.83423\n",
      "\t15: \"SibSp\" NUMERICAL mean:0.524184 min:0 max:8 sd:1.10308\n",
      "\n",
      "CATEGORICAL: 7 (35%)\n",
      "\t1: \"Cabin\" CATEGORICAL num-nas:687 (77.2778%) has-dict vocab-size:1 num-oods:146 (72.2772%)\n",
      "\t2: \"Embarked\" CATEGORICAL has-dict vocab-size:4 zero-ood-items most-frequent:\"S\" 644 (72.4409%)\n",
      "\t14: \"Sex\" CATEGORICAL has-dict vocab-size:3 zero-ood-items most-frequent:\"male\" 577 (64.9044%)\n",
      "\t16: \"Ticket\" CATEGORICAL has-dict vocab-size:9 num-oods:672 (75.5906%) most-frequent:\"<OOD>\" 672 (75.5906%)\n",
      "\t17: \"Ticket_item\" CATEGORICAL has-dict vocab-size:17 num-oods:28 (3.14961%) most-frequent:\"None\" 663 (74.5782%)\n",
      "\t18: \"Ticket_number\" CATEGORICAL has-dict vocab-size:9 num-oods:670 (75.3656%) most-frequent:\"<OOD>\" 670 (75.3656%)\n",
      "\t19: \"__LABEL\" CATEGORICAL integerized vocab-size:3 no-ood-item\n",
      "\n",
      "Terminology:\n",
      "\tnas: Number of non-available (i.e. missing) values.\n",
      "\tood: Out of dictionary.\n",
      "\tmanually-defined: Attribute which type is manually defined by the user i.e. the type was not automatically inferred.\n",
      "\ttokenized: The attribute value is obtained through tokenization.\n",
      "\thas-dict: The attribute is attached to a string dictionary e.g. a categorical attribute stored as a string.\n",
      "\tvocab-size: Number of unique values.\n",
      "\n",
      "[INFO 2023-05-30T18:21:18.037994701+00:00 kernel.cc:790] Configure learner\n",
      "[INFO 2023-05-30T18:21:18.038356233+00:00 kernel.cc:804] Training config:\n",
      "learner: \"HYPERPARAMETER_OPTIMIZER\"\n",
      "features: \"^Age$\"\n",
      "features: \"^Cabin$\"\n",
      "features: \"^Embarked$\"\n",
      "features: \"^Fare$\"\n",
      "features: \"^PC1$\"\n",
      "features: \"^PC2$\"\n",
      "features: \"^PC3$\"\n",
      "features: \"^PC4$\"\n",
      "features: \"^PC5$\"\n",
      "features: \"^PC6$\"\n",
      "features: \"^PC7$\"\n",
      "features: \"^Parch$\"\n",
      "features: \"^PassengerId$\"\n",
      "features: \"^Pclass$\"\n",
      "features: \"^Sex$\"\n",
      "features: \"^SibSp$\"\n",
      "features: \"^Ticket$\"\n",
      "features: \"^Ticket_item$\"\n",
      "features: \"^Ticket_number$\"\n",
      "label: \"^__LABEL$\"\n",
      "task: CLASSIFICATION\n",
      "metadata {\n",
      "  framework: \"TF Keras\"\n",
      "}\n",
      "[yggdrasil_decision_forests.model.hyperparameters_optimizer_v2.proto.hyperparameters_optimizer_config] {\n",
      "  base_learner {\n",
      "    learner: \"GRADIENT_BOOSTED_TREES\"\n",
      "    features: \"^Age$\"\n",
      "    features: \"^Cabin$\"\n",
      "    features: \"^Embarked$\"\n",
      "    features: \"^Fare$\"\n",
      "    features: \"^PC1$\"\n",
      "    features: \"^PC2$\"\n",
      "    features: \"^PC3$\"\n",
      "    features: \"^PC4$\"\n",
      "    features: \"^PC5$\"\n",
      "    features: \"^PC6$\"\n",
      "    features: \"^PC7$\"\n",
      "    features: \"^Parch$\"\n",
      "    features: \"^PassengerId$\"\n",
      "    features: \"^Pclass$\"\n",
      "    features: \"^Sex$\"\n",
      "    features: \"^SibSp$\"\n",
      "    features: \"^Ticket$\"\n",
      "    features: \"^Ticket_item$\"\n",
      "    features: \"^Ticket_number$\"\n",
      "    label: \"^__LABEL$\"\n",
      "    task: CLASSIFICATION\n",
      "    random_seed: 123456\n",
      "    pure_serving_model: false\n",
      "    [yggdrasil_decision_forests.model.gradient_boosted_trees.proto.gradient_boosted_trees_config] {\n",
      "      num_trees: 300\n",
      "      decision_tree {\n",
      "        max_depth: 6\n",
      "        min_examples: 5\n",
      "        in_split_min_examples_check: true\n",
      "        keep_non_leaf_label_distribution: true\n",
      "        num_candidate_attributes: -1\n",
      "        missing_value_policy: GLOBAL_IMPUTATION\n",
      "        allow_na_conditions: false\n",
      "        categorical_set_greedy_forward {\n",
      "          sampling: 0.1\n",
      "          max_num_items: -1\n",
      "          min_item_frequency: 1\n",
      "        }\n",
      "        growing_strategy_local {\n",
      "        }\n",
      "        categorical {\n",
      "          cart {\n",
      "          }\n",
      "        }\n",
      "        axis_aligned_split {\n",
      "        }\n",
      "        internal {\n",
      "          sorting_strategy: PRESORTED\n",
      "        }\n",
      "        uplift {\n",
      "          min_examples_in_treatment: 5\n",
      "          split_score: KULLBACK_LEIBLER\n",
      "        }\n",
      "      }\n",
      "      shrinkage: 0.1\n",
      "      loss: DEFAULT\n",
      "      validation_set_ratio: 0.1\n",
      "      validation_interval_in_trees: 1\n",
      "      early_stopping: VALIDATION_LOSS_INCREASE\n",
      "      early_stopping_num_trees_look_ahead: 30\n",
      "      l2_regularization: 0\n",
      "      lambda_loss: 1\n",
      "      mart {\n",
      "      }\n",
      "      adapt_subsample_for_maximum_training_duration: false\n",
      "      l1_regularization: 0\n",
      "      use_hessian_gain: false\n",
      "      l2_regularization_categorical: 1\n",
      "      stochastic_gradient_boosting {\n",
      "        ratio: 1\n",
      "      }\n",
      "      apply_link_function: true\n",
      "      compute_permutation_variable_importance: false\n",
      "      binary_focal_loss_options {\n",
      "        misprediction_exponent: 2\n",
      "        positive_sample_coefficient: 0.5\n",
      "      }\n",
      "      early_stopping_initial_iteration: 10\n",
      "    }\n",
      "  }\n",
      "  optimizer {\n",
      "    optimizer_key: \"RANDOM\"\n",
      "    [yggdrasil_decision_forests.model.hyperparameters_optimizer_v2.proto.random] {\n",
      "      num_trials: 100\n",
      "    }\n",
      "  }\n",
      "  base_learner_deployment {\n",
      "    num_threads: 1\n",
      "  }\n",
      "  predefined_search_space {\n",
      "  }\n",
      "}\n",
      "\n",
      "[INFO 2023-05-30T18:21:18.03862849+00:00 kernel.cc:807] Deployment config:\n",
      "cache_path: \"/tmp/tmplf_8vqpz/working_cache\"\n",
      "num_threads: 4\n",
      "try_resume_training: true\n",
      "\n",
      "[INFO 2023-05-30T18:21:18.038879498+00:00 kernel.cc:868] Train model\n",
      "[INFO 2023-05-30T18:21:20.947204497+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.997237\n",
      "[INFO 2023-05-30T18:21:21.439993537+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.990506\n",
      "[INFO 2023-05-30T18:21:21.453382779+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.943168\n",
      "[INFO 2023-05-30T18:21:24.171410295+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.955766\n",
      "[INFO 2023-05-30T18:21:24.241426393+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.992669\n",
      "[INFO 2023-05-30T18:21:24.950627152+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.992061\n",
      "[INFO 2023-05-30T18:21:25.052730985+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.00775\n",
      "[INFO 2023-05-30T18:21:25.579499842+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.987968\n",
      "[INFO 2023-05-30T18:21:26.657539157+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.00342\n",
      "[INFO 2023-05-30T18:21:28.703886176+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.03234\n",
      "[INFO 2023-05-30T18:21:28.796843022+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.941193\n",
      "[INFO 2023-05-30T18:21:30.477104443+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.947625\n",
      "[INFO 2023-05-30T18:21:30.924126282+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.979337\n",
      "[INFO 2023-05-30T18:21:33.434838071+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.981829\n",
      "[INFO 2023-05-30T18:21:34.333924602+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.996775\n",
      "[INFO 2023-05-30T18:21:34.970150054+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.993219\n",
      "[INFO 2023-05-30T18:21:35.148404527+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.996244\n",
      "[INFO 2023-05-30T18:21:36.452311449+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.939293\n",
      "[INFO 2023-05-30T18:21:37.842571378+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.999846\n",
      "[INFO 2023-05-30T18:21:40.537696647+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.983169\n",
      "[INFO 2023-05-30T18:21:40.845018472+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.944815\n",
      "[INFO 2023-05-30T18:21:42.445258397+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.967603\n",
      "[INFO 2023-05-30T18:21:42.766446203+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.930141\n",
      "[INFO 2023-05-30T18:21:44.301863451+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.934558\n",
      "[INFO 2023-05-30T18:21:45.273722499+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.953716\n",
      "[INFO 2023-05-30T18:21:46.512169223+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.996593\n",
      "[INFO 2023-05-30T18:21:49.04851006+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.998171\n",
      "[INFO 2023-05-30T18:21:50.775478635+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.983638\n",
      "[INFO 2023-05-30T18:21:51.156305393+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.977461\n",
      "[INFO 2023-05-30T18:21:51.34814847+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.00811\n",
      "[INFO 2023-05-30T18:21:53.009018764+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.02599\n",
      "[INFO 2023-05-30T18:21:53.403155718+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.997956\n",
      "[INFO 2023-05-30T18:21:56.699069396+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.02811\n",
      "[INFO 2023-05-30T18:21:56.720508387+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.983629\n",
      "[INFO 2023-05-30T18:21:57.346261138+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.02334\n",
      "[INFO 2023-05-30T18:21:58.173183505+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.977457\n",
      "[INFO 2023-05-30T18:21:58.210912253+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.00996\n",
      "[INFO 2023-05-30T18:21:58.435068237+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.972909\n",
      "[INFO 2023-05-30T18:22:00.52209492+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.996843\n",
      "[INFO 2023-05-30T18:22:00.849060532+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.995948\n",
      "[INFO 2023-05-30T18:22:02.387134383+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.933925\n",
      "[INFO 2023-05-30T18:22:03.159474091+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.948491\n",
      "[INFO 2023-05-30T18:22:03.779706995+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.989515\n",
      "[INFO 2023-05-30T18:22:04.435168676+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.994856\n",
      "[INFO 2023-05-30T18:22:06.621211691+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.988891\n",
      "[INFO 2023-05-30T18:22:07.020843688+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.994002\n",
      "[INFO 2023-05-30T18:22:08.301869052+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.983659\n",
      "[INFO 2023-05-30T18:22:09.732695082+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.993159\n",
      "[INFO 2023-05-30T18:22:11.24121709+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.02703\n",
      "[INFO 2023-05-30T18:22:11.282402574+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.970428\n",
      "[INFO 2023-05-30T18:22:11.718156983+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.968105\n",
      "[INFO 2023-05-30T18:22:11.789680237+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.975162\n",
      "[INFO 2023-05-30T18:22:12.449675809+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.00774\n",
      "[INFO 2023-05-30T18:22:12.778225492+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.978579\n",
      "[INFO 2023-05-30T18:22:14.109510061+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.9754\n",
      "[INFO 2023-05-30T18:22:15.702615897+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.979191\n",
      "[INFO 2023-05-30T18:22:15.883416823+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.01147\n",
      "[INFO 2023-05-30T18:22:16.544198254+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.981374\n",
      "[INFO 2023-05-30T18:22:18.142384887+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.977212\n",
      "[INFO 2023-05-30T18:22:19.290152427+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.95637\n",
      "[INFO 2023-05-30T18:22:20.538542705+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.988656\n",
      "[INFO 2023-05-30T18:22:20.543497672+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.971337\n",
      "[INFO 2023-05-30T18:22:21.493036657+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.993267\n",
      "[INFO 2023-05-30T18:22:22.279778403+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.983818\n",
      "[INFO 2023-05-30T18:22:23.293108673+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.947222\n",
      "[INFO 2023-05-30T18:22:24.373969617+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.0062\n",
      "[INFO 2023-05-30T18:22:24.702207228+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.993888\n",
      "[INFO 2023-05-30T18:22:24.759562286+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.994036\n",
      "[INFO 2023-05-30T18:22:27.693009141+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.01544\n",
      "[INFO 2023-05-30T18:22:28.662813609+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.967202\n",
      "[INFO 2023-05-30T18:22:29.428114239+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.975984\n",
      "[INFO 2023-05-30T18:22:30.016246183+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.993456\n",
      "[INFO 2023-05-30T18:22:30.421192377+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.950909\n",
      "[INFO 2023-05-30T18:22:32.163738072+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.968506\n",
      "[INFO 2023-05-30T18:22:32.519115339+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.927769\n",
      "[INFO 2023-05-30T18:22:33.994633004+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.951556\n",
      "[INFO 2023-05-30T18:22:34.452201718+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.97372\n",
      "[INFO 2023-05-30T18:22:35.010329417+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.974169\n",
      "[INFO 2023-05-30T18:22:35.856922454+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.980981\n",
      "[INFO 2023-05-30T18:22:38.226792759+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.955825\n",
      "[INFO 2023-05-30T18:22:38.574883873+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.978434\n",
      "[INFO 2023-05-30T18:22:40.529730733+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.04817\n",
      "[INFO 2023-05-30T18:22:41.630479309+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.958687\n",
      "[INFO 2023-05-30T18:22:41.70318501+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.02373\n",
      "[INFO 2023-05-30T18:22:44.302769557+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.955736\n",
      "[INFO 2023-05-30T18:22:45.322667202+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.972394\n",
      "[INFO 2023-05-30T18:22:45.867092623+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.02718\n",
      "[INFO 2023-05-30T18:22:48.151415171+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.987202\n",
      "[INFO 2023-05-30T18:22:49.065310806+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.969939\n",
      "[INFO 2023-05-30T18:22:51.249113511+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.0156\n",
      "[INFO 2023-05-30T18:22:51.532351547+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.972442\n",
      "[INFO 2023-05-30T18:22:54.474707966+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.01694\n",
      "[INFO 2023-05-30T18:22:55.960127948+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 1.00591\n",
      "[INFO 2023-05-30T18:22:57.471849304+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.991875\n",
      "[INFO 2023-05-30T18:22:58.576585685+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.981803\n",
      "[INFO 2023-05-30T18:22:59.639550581+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.987777\n",
      "[INFO 2023-05-30T18:22:59.702471499+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.963883\n",
      "[INFO 2023-05-30T18:23:00.286251753+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.98672\n",
      "[INFO 2023-05-30T18:23:01.146846092+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.979974\n",
      "[INFO 2023-05-30T18:23:03.348676042+00:00 early_stopping.cc:53] Early stop of the training because the validation loss does not decrease anymore. Best valid-loss: 0.992669\n",
      "[INFO 2023-05-30T18:23:03.357410093+00:00 kernel.cc:905] Export model in log directory: /tmp/tmplf_8vqpz with prefix 4aeb80bd481749ed\n",
      "[INFO 2023-05-30T18:23:03.365620365+00:00 kernel.cc:923] Save model in resources\n",
      "[INFO 2023-05-30T18:23:03.37464187+00:00 abstract_model.cc:849] Model self evaluation:\n",
      "Task: CLASSIFICATION\n",
      "Label: __LABEL\n",
      "Loss (BINOMIAL_LOG_LIKELIHOOD): 0.927769\n",
      "\n",
      "Accuracy: 0.794521  CI95[W][0 1]\n",
      "ErrorRate: : 0.205479\n",
      "\n",
      "\n",
      "Confusion Table:\n",
      "truth\\prediction\n",
      "   0   1   2\n",
      "0  0   0   0\n",
      "1  0  37   4\n",
      "2  0  11  21\n",
      "Total: 73\n",
      "\n",
      "One vs other classes:\n",
      "\n",
      "[INFO 2023-05-30T18:23:03.422557811+00:00 kernel.cc:1214] Loading model from path /tmp/tmplf_8vqpz/model/ with prefix 4aeb80bd481749ed\n",
      "[INFO 2023-05-30T18:23:03.442301602+00:00 decision_forest.cc:661] Model loaded with 29 root(s), 1277 node(s), and 18 input feature(s).\n",
      "[INFO 2023-05-30T18:23:03.442427308+00:00 abstract_model.cc:1311] Engine \"GradientBoostedTreesGeneric\" built\n",
      "[INFO 2023-05-30T18:23:03.442518584+00:00 kernel.cc:1046] Use fast generic engine\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model trained in 0:01:45.429603\n",
      "Compiling model...\n",
      "Model compiled.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7e6190dffa90>"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# train the model with automatic tuning\n",
    "auto_tuned_model = tfdf.keras.GradientBoostedTreesModel(tuner = tuner)\n",
    "auto_tuned_model.fit(train_ds, verbose = 2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "ba14c5c9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:23:03.954946Z",
     "iopub.status.busy": "2023-05-30T18:23:03.953529Z",
     "iopub.status.idle": "2023-05-30T18:23:03.967040Z",
     "shell.execute_reply": "2023-05-30T18:23:03.965817Z"
    },
    "papermill": {
     "duration": 0.067015,
     "end_time": "2023-05-30T18:23:03.969549",
     "exception": false,
     "start_time": "2023-05-30T18:23:03.902534",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.7945205569267273    Loss: 0.927769124507904\n"
     ]
    }
   ],
   "source": [
    "self_evaluation = auto_tuned_model.make_inspector().evaluation()\n",
    "print(f\"Accuracy: {self_evaluation.accuracy}    Loss: {self_evaluation.loss}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "dfd6beb1",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:23:04.070037Z",
     "iopub.status.busy": "2023-05-30T18:23:04.069234Z",
     "iopub.status.idle": "2023-05-30T18:23:04.302950Z",
     "shell.execute_reply": "2023-05-30T18:23:04.301999Z"
    },
    "papermill": {
     "duration": 0.286493,
     "end_time": "2023-05-30T18:23:04.305476",
     "exception": false,
     "start_time": "2023-05-30T18:23:04.018983",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9190101027488708"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# valuate the model\n",
    "auto_tuned_model.compile(['accuracy'])\n",
    "model_accuracy = auto_tuned_model.evaluate(train_ds, return_dict = True, verbose = 0)['accuracy']\n",
    "model_accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "221e56dd",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:23:04.404828Z",
     "iopub.status.busy": "2023-05-30T18:23:04.403681Z",
     "iopub.status.idle": "2023-05-30T18:23:04.426965Z",
     "shell.execute_reply": "2023-05-30T18:23:04.425646Z"
    },
    "papermill": {
     "duration": 0.075497,
     "end_time": "2023-05-30T18:23:04.429269",
     "exception": false,
     "start_time": "2023-05-30T18:23:04.353772",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(100, 17)\n"
     ]
    }
   ],
   "source": [
    "# The hyper-parameters and objective scores of the trials are available in the model inspector. \n",
    "# display the tuning logs\n",
    "tuning_logs = auto_tuned_model.make_inspector().tuning_logs()\n",
    "print(tuning_logs.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "dccbce56",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:23:04.528247Z",
     "iopub.status.busy": "2023-05-30T18:23:04.527683Z",
     "iopub.status.idle": "2023-05-30T18:23:04.556987Z",
     "shell.execute_reply": "2023-05-30T18:23:04.555605Z"
    },
    "papermill": {
     "duration": 0.081943,
     "end_time": "2023-05-30T18:23:04.559408",
     "exception": false,
     "start_time": "2023-05-30T18:23:04.477465",
     "status": "completed"
    },
    "tags": []
   },
   "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>score</th>\n",
       "      <th>evaluation_time</th>\n",
       "      <th>best</th>\n",
       "      <th>split_axis</th>\n",
       "      <th>sparse_oblique_projection_density_factor</th>\n",
       "      <th>sparse_oblique_normalization</th>\n",
       "      <th>sparse_oblique_weights</th>\n",
       "      <th>categorical_algorithm</th>\n",
       "      <th>growing_strategy</th>\n",
       "      <th>max_num_nodes</th>\n",
       "      <th>sampling_method</th>\n",
       "      <th>subsample</th>\n",
       "      <th>shrinkage</th>\n",
       "      <th>min_examples</th>\n",
       "      <th>use_hessian_gain</th>\n",
       "      <th>num_candidate_attributes_ratio</th>\n",
       "      <th>max_depth</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.997237</td>\n",
       "      <td>2.912175</td>\n",
       "      <td>False</td>\n",
       "      <td>SPARSE_OBLIQUE</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NONE</td>\n",
       "      <td>BINARY</td>\n",
       "      <td>CART</td>\n",
       "      <td>BEST_FIRST_GLOBAL</td>\n",
       "      <td>64.0</td>\n",
       "      <td>RANDOM</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>20</td>\n",
       "      <td>false</td>\n",
       "      <td>0.2</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.990506</td>\n",
       "      <td>3.401996</td>\n",
       "      <td>False</td>\n",
       "      <td>SPARSE_OBLIQUE</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NONE</td>\n",
       "      <td>BINARY</td>\n",
       "      <td>RANDOM</td>\n",
       "      <td>LOCAL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>RANDOM</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>5</td>\n",
       "      <td>false</td>\n",
       "      <td>0.5</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.943168</td>\n",
       "      <td>3.414845</td>\n",
       "      <td>False</td>\n",
       "      <td>SPARSE_OBLIQUE</td>\n",
       "      <td>5.0</td>\n",
       "      <td>MIN_MAX</td>\n",
       "      <td>BINARY</td>\n",
       "      <td>CART</td>\n",
       "      <td>BEST_FIRST_GLOBAL</td>\n",
       "      <td>32.0</td>\n",
       "      <td>RANDOM</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>10</td>\n",
       "      <td>false</td>\n",
       "      <td>0.2</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.955766</td>\n",
       "      <td>6.132722</td>\n",
       "      <td>False</td>\n",
       "      <td>SPARSE_OBLIQUE</td>\n",
       "      <td>4.0</td>\n",
       "      <td>STANDARD_DEVIATION</td>\n",
       "      <td>CONTINUOUS</td>\n",
       "      <td>RANDOM</td>\n",
       "      <td>BEST_FIRST_GLOBAL</td>\n",
       "      <td>256.0</td>\n",
       "      <td>RANDOM</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>20</td>\n",
       "      <td>true</td>\n",
       "      <td>0.5</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.992669</td>\n",
       "      <td>6.207620</td>\n",
       "      <td>False</td>\n",
       "      <td>SPARSE_OBLIQUE</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NONE</td>\n",
       "      <td>BINARY</td>\n",
       "      <td>RANDOM</td>\n",
       "      <td>BEST_FIRST_GLOBAL</td>\n",
       "      <td>128.0</td>\n",
       "      <td>RANDOM</td>\n",
       "      <td>0.8</td>\n",
       "      <td>0.02</td>\n",
       "      <td>10</td>\n",
       "      <td>true</td>\n",
       "      <td>0.9</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      score  evaluation_time   best      split_axis  \\\n",
       "0 -0.997237         2.912175  False  SPARSE_OBLIQUE   \n",
       "1 -0.990506         3.401996  False  SPARSE_OBLIQUE   \n",
       "2 -0.943168         3.414845  False  SPARSE_OBLIQUE   \n",
       "3 -0.955766         6.132722  False  SPARSE_OBLIQUE   \n",
       "4 -0.992669         6.207620  False  SPARSE_OBLIQUE   \n",
       "\n",
       "   sparse_oblique_projection_density_factor sparse_oblique_normalization  \\\n",
       "0                                       1.0                         NONE   \n",
       "1                                       3.0                         NONE   \n",
       "2                                       5.0                      MIN_MAX   \n",
       "3                                       4.0           STANDARD_DEVIATION   \n",
       "4                                       2.0                         NONE   \n",
       "\n",
       "  sparse_oblique_weights categorical_algorithm   growing_strategy  \\\n",
       "0                 BINARY                  CART  BEST_FIRST_GLOBAL   \n",
       "1                 BINARY                RANDOM              LOCAL   \n",
       "2                 BINARY                  CART  BEST_FIRST_GLOBAL   \n",
       "3             CONTINUOUS                RANDOM  BEST_FIRST_GLOBAL   \n",
       "4                 BINARY                RANDOM  BEST_FIRST_GLOBAL   \n",
       "\n",
       "   max_num_nodes sampling_method  subsample  shrinkage  min_examples  \\\n",
       "0           64.0          RANDOM        1.0       0.10            20   \n",
       "1            NaN          RANDOM        1.0       0.10             5   \n",
       "2           32.0          RANDOM        1.0       0.10            10   \n",
       "3          256.0          RANDOM        1.0       0.10            20   \n",
       "4          128.0          RANDOM        0.8       0.02            10   \n",
       "\n",
       "  use_hessian_gain  num_candidate_attributes_ratio  max_depth  \n",
       "0            false                             0.2        NaN  \n",
       "1            false                             0.5        8.0  \n",
       "2            false                             0.2        NaN  \n",
       "3             true                             0.5        NaN  \n",
       "4             true                             0.9        NaN  "
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The score value is always maximized. \n",
    "# score below is the negative log loss on the validation dataset (selected automatically).\n",
    "tuning_logs.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "3877049d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:23:04.658706Z",
     "iopub.status.busy": "2023-05-30T18:23:04.658245Z",
     "iopub.status.idle": "2023-05-30T18:23:04.670819Z",
     "shell.execute_reply": "2023-05-30T18:23:04.669670Z"
    },
    "papermill": {
     "duration": 0.065036,
     "end_time": "2023-05-30T18:23:04.673099",
     "exception": false,
     "start_time": "2023-05-30T18:23:04.608063",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "score                                               -0.927769\n",
       "evaluation_time                                     74.480449\n",
       "best                                                     True\n",
       "split_axis                                     SPARSE_OBLIQUE\n",
       "sparse_oblique_projection_density_factor                  1.0\n",
       "sparse_oblique_normalization                             NONE\n",
       "sparse_oblique_weights                             CONTINUOUS\n",
       "categorical_algorithm                                  RANDOM\n",
       "growing_strategy                            BEST_FIRST_GLOBAL\n",
       "max_num_nodes                                           512.0\n",
       "sampling_method                                        RANDOM\n",
       "subsample                                                 0.6\n",
       "shrinkage                                                 0.1\n",
       "min_examples                                                7\n",
       "use_hessian_gain                                        false\n",
       "num_candidate_attributes_ratio                            0.5\n",
       "max_depth                                                 NaN\n",
       "Name: 74, dtype: object"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Best hyper-parameters.\n",
    "tuning_logs[tuning_logs.best].iloc[0]\n",
    "\n",
    "# Parameters with NaN value are conditional parameters that are not set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "bb307041",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:23:04.772877Z",
     "iopub.status.busy": "2023-05-30T18:23:04.772249Z",
     "iopub.status.idle": "2023-05-30T18:23:05.143725Z",
     "shell.execute_reply": "2023-05-30T18:23:05.142487Z"
    },
    "papermill": {
     "duration": 0.42438,
     "end_time": "2023-05-30T18:23:05.146546",
     "exception": false,
     "start_time": "2023-05-30T18:23:04.722166",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot the evaluation of the best score during tuning\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.figure(figsize=(10, 5))\n",
    "plt.plot(tuning_logs[\"score\"], label=\"current trial\")\n",
    "plt.plot(tuning_logs[\"score\"].cummax(), label=\"best trial\")\n",
    "plt.xlabel(\"Tuning step\")\n",
    "plt.ylabel(\"Tuning score\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1e286968",
   "metadata": {
    "papermill": {
     "duration": 0.050129,
     "end_time": "2023-05-30T18:23:05.247121",
     "exception": false,
     "start_time": "2023-05-30T18:23:05.196992",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# **4. Submission**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "df6bb743",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:23:05.352246Z",
     "iopub.status.busy": "2023-05-30T18:23:05.350976Z",
     "iopub.status.idle": "2023-05-30T18:23:05.459422Z",
     "shell.execute_reply": "2023-05-30T18:23:05.458158Z"
    },
    "papermill": {
     "duration": 0.164172,
     "end_time": "2023-05-30T18:23:05.462307",
     "exception": false,
     "start_time": "2023-05-30T18:23:05.298135",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 82ms/step\n"
     ]
    }
   ],
   "source": [
    "# create output datafram and save to a csv file\n",
    "def prediction(model, threshold = 0.5):\n",
    "    proba = model.predict(test_ds)[:,0]\n",
    "    return pd.DataFrame({'PassengerID': pre1_test_data['PassengerId'], 'Survived': (proba>=threshold).astype(int)})\n",
    "\n",
    "def make_summision(yhat_df):\n",
    "    return yhat_df.to_csv(\"submission.csv\", index = False)\n",
    "    print('Submission was successful!')\n",
    "\n",
    "yhat_df = prediction(auto_tuned_model)\n",
    "make_summision(yhat_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "5f2c51e9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-05-30T18:23:05.566846Z",
     "iopub.status.busy": "2023-05-30T18:23:05.565652Z",
     "iopub.status.idle": "2023-05-30T18:23:05.578549Z",
     "shell.execute_reply": "2023-05-30T18:23:05.577384Z"
    },
    "papermill": {
     "duration": 0.067143,
     "end_time": "2023-05-30T18:23:05.580972",
     "exception": false,
     "start_time": "2023-05-30T18:23:05.513829",
     "status": "completed"
    },
    "tags": []
   },
   "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>892</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>893</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>894</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>895</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>896</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>413</th>\n",
       "      <td>1305</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>414</th>\n",
       "      <td>1306</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>415</th>\n",
       "      <td>1307</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>416</th>\n",
       "      <td>1308</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>417</th>\n",
       "      <td>1309</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>418 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     PassengerID  Survived\n",
       "0            892         0\n",
       "1            893         0\n",
       "2            894         0\n",
       "3            895         0\n",
       "4            896         0\n",
       "..           ...       ...\n",
       "413         1305         0\n",
       "414         1306         1\n",
       "415         1307         0\n",
       "416         1308         0\n",
       "417         1309         0\n",
       "\n",
       "[418 rows x 2 columns]"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# preview the predictions on test data\n",
    "yhat_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c5bfbf73",
   "metadata": {
    "papermill": {
     "duration": 0.050812,
     "end_time": "2023-05-30T18:23:05.682478",
     "exception": false,
     "start_time": "2023-05-30T18:23:05.631666",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.10.10"
  },
  "papermill": {
   "default_parameters": {},
   "duration": 154.486343,
   "end_time": "2023-05-30T18:23:09.295702",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2023-05-30T18:20:34.809359",
   "version": "2.4.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}