HyeongWookKim/Kaggle_Titanic_Tutorial_Part2.ipynb

## Kaggle_Titanic_Tutorial_Part2.ipynb
{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Kaggle - Titanic Tutorial_Part 2\n",
    "**<분석 목적>**\n",
    "- 타이타닉에 탑승한 사람들의 신상정보를 활용하여, 승선한 사람들의 **생존 여부**를 **예측**하는 모델을 생성하는 것이다.\n",
    "\n",
    "**<튜토리얼 과정>**\n",
    "1. 데이터 셋 확인\n",
    "    - Null 값 존재 여부를 확인하고, Null 값들에 대한 처리를 해준다.\n",
    "2. 탐색적 데이터 분석(EDA)\n",
    "    - 여러 개의 feature들을 개별적으로 분석하고, feature들 간의 상관관계를 확인한다.\n",
    "    - 또한 시각화 라이브러리들을 활용하여 insight를 도출한다.\n",
    "3. Feature Engineering(특성 공학)\n",
    "    - 모델 설정에 앞서, 모델의 성능을 높이기 위한 작업이다.\n",
    "    - 원-핫-인코딩(One-hot-encoding), class 나누기, 구간 나누기, 텍스트 데이터 처리 등...\n",
    "4. 모델 생성\n",
    "    - 사이킷런을 사용해서 모델을 만든다.\n",
    "5. 모델 학습 및 예측\n",
    "    - 학습 데이터 셋(train dataset)을 가지고 모델을 학습시킨 후, 테스트 데이터 셋(test dataset)을 가지고 예측을 수행한다.\n",
    "6. 모델 평가\n",
    "    - 모델의 예측 성능을 평가한다.\n",
    "    - 우리가 직면한 문제에 맞게 모델 평가 지표를 결정하면 된다.\n",
    "        - ex) RMSE, R_square, F1-score, ..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:37.378244Z",
     "start_time": "2020-06-09T06:37:35.860463Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "        <script type=\"text/javascript\">\n",
       "        window.PlotlyConfig = {MathJaxConfig: 'local'};\n",
       "        if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}\n",
       "        if (typeof require !== 'undefined') {\n",
       "        require.undef(\"plotly\");\n",
       "        requirejs.config({\n",
       "            paths: {\n",
       "                'plotly': ['https://cdn.plot.ly/plotly-latest.min']\n",
       "            }\n",
       "        });\n",
       "        require(['plotly'], function(Plotly) {\n",
       "            window._Plotly = Plotly;\n",
       "        });\n",
       "        }\n",
       "        </script>\n",
       "        "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from pandas import Series\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "%matplotlib inline\n",
    "\n",
    "# 이 두 줄의 코드는 matplotlib의 기본 scheme말고, seaborn scheme을 세팅해준다\n",
    "# 일일이 graph의 font size를 지정할 필요 없이, seaborn의 font_scale을 사용하면 편리하다\n",
    "plt.style.use('seaborn')\n",
    "sns.set(font_scale = 2.5)\n",
    "\n",
    "import plotly.offline as py\n",
    "py.init_notebook_mode(connected = True)\n",
    "import plotly.graph_objs as go\n",
    "import plotly.tools as tls\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**본격적으로 <Part 2>를 시작하기에 앞서, <Part 1>에서 수행해주었던 작업들을 다시 수행해주겠다.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:37.430117Z",
     "start_time": "2020-06-09T06:37:37.381236Z"
    }
   },
   "outputs": [],
   "source": [
    "df_train = pd.read_csv('../titanic/train.csv')\n",
    "df_test = pd.read_csv('../titanic/test.csv')\n",
    "\n",
    "# SibSp, Parch 변수를 하나의 변수(FamilySize)로 합쳐준다\n",
    "# 자신을 포함해야하므로, 1 을 더해준다\n",
    "df_train['FamilySize'] = df_train['SibSp'] + df_train['Parch'] + 1\n",
    "df_test['FamilySize'] = df_test['SibSp'] + df_test['Parch'] + 1\n",
    "\n",
    "# Fare 변수의 결측치들을 평균값으로 대체\n",
    "df_test.loc[df_test.Fare.isnull(), 'Fare'] = df_test['Fare'].mean()\n",
    "\n",
    "# Fare 변수에 대해서 '로그 변환' 수행\n",
    "df_train['Fare'] = df_train['Fare'].map(lambda x: np.log(x) if x > 0 else 0)\n",
    "df_test['Fare'] = df_test['Fare'].map(lambda x: np.log(x) if x > 0 else 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature Engineering(특성 공학)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Null 값 처리"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**1. Age 변수의 Null 값 처리**\n",
    "- Age에는 Null 값이 177개나 존재한다.\n",
    "- 이를 채워주기 위해, **title + statistics** 를 사용해보겠다.\n",
    "    - 영어에서는 Mr, Mrs, Miss와 같은 title이 존재한다.\n",
    "    - 때문에 각 탑승객의 이름에는 위와 같은 title이 들어가게 되는데, 이를 사용해보겠다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:37.447062Z",
     "start_time": "2020-06-09T06:37:37.433097Z"
    }
   },
   "outputs": [],
   "source": [
    "# 정규표현식을 사용하여 title을 추출\n",
    "df_train['Initial'] = df_train.Name.str.extract('([A-Za-z]+)\\.')\n",
    "df_test['Initial'] = df_test.Name.str.extract('([A-Za-z]+)\\.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:37.822594Z",
     "start_time": "2020-06-09T06:37:37.450053Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col0 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col1 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col2 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col3 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col4 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col5 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col6 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col7 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col8 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col9 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col10 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col11 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col12 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col13 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col14 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col15 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col16 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col0 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col1 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col2 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col3 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col4 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col5 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col6 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col7 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col8 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col9 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col10 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col11 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col12 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col13 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col14 {\n",
       "            background-color:  #ffff66;\n",
       "            color:  #000000;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col15 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }    #T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col16 {\n",
       "            background-color:  #008066;\n",
       "            color:  #f1f1f1;\n",
       "        }</style><table id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60\" ><thead>    <tr>        <th class=\"index_name level0\" >Initial</th>        <th class=\"col_heading level0 col0\" >Capt</th>        <th class=\"col_heading level0 col1\" >Col</th>        <th class=\"col_heading level0 col2\" >Countess</th>        <th class=\"col_heading level0 col3\" >Don</th>        <th class=\"col_heading level0 col4\" >Dr</th>        <th class=\"col_heading level0 col5\" >Jonkheer</th>        <th class=\"col_heading level0 col6\" >Lady</th>        <th class=\"col_heading level0 col7\" >Major</th>        <th class=\"col_heading level0 col8\" >Master</th>        <th class=\"col_heading level0 col9\" >Miss</th>        <th class=\"col_heading level0 col10\" >Mlle</th>        <th class=\"col_heading level0 col11\" >Mme</th>        <th class=\"col_heading level0 col12\" >Mr</th>        <th class=\"col_heading level0 col13\" >Mrs</th>        <th class=\"col_heading level0 col14\" >Ms</th>        <th class=\"col_heading level0 col15\" >Rev</th>        <th class=\"col_heading level0 col16\" >Sir</th>    </tr>    <tr>        <th class=\"index_name level0\" >Sex</th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60level0_row0\" class=\"row_heading level0 row0\" >female</th>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col0\" class=\"data row0 col0\" >0</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col1\" class=\"data row0 col1\" >0</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col2\" class=\"data row0 col2\" >1</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col3\" class=\"data row0 col3\" >0</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col4\" class=\"data row0 col4\" >1</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col5\" class=\"data row0 col5\" >0</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col6\" class=\"data row0 col6\" >1</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col7\" class=\"data row0 col7\" >0</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col8\" class=\"data row0 col8\" >0</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col9\" class=\"data row0 col9\" >182</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col10\" class=\"data row0 col10\" >2</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col11\" class=\"data row0 col11\" >1</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col12\" class=\"data row0 col12\" >0</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col13\" class=\"data row0 col13\" >125</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col14\" class=\"data row0 col14\" >1</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col15\" class=\"data row0 col15\" >0</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row0_col16\" class=\"data row0 col16\" >0</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60level0_row1\" class=\"row_heading level0 row1\" >male</th>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col0\" class=\"data row1 col0\" >1</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col1\" class=\"data row1 col1\" >2</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col2\" class=\"data row1 col2\" >0</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col3\" class=\"data row1 col3\" >1</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col4\" class=\"data row1 col4\" >6</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col5\" class=\"data row1 col5\" >1</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col6\" class=\"data row1 col6\" >0</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col7\" class=\"data row1 col7\" >2</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col8\" class=\"data row1 col8\" >40</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col9\" class=\"data row1 col9\" >0</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col10\" class=\"data row1 col10\" >0</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col11\" class=\"data row1 col11\" >0</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col12\" class=\"data row1 col12\" >517</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col13\" class=\"data row1 col13\" >0</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col14\" class=\"data row1 col14\" >0</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col15\" class=\"data row1 col15\" >6</td>\n",
       "                        <td id=\"T_b5bd59f6_aa1b_11ea_8a3b_9822ef754c60row1_col16\" class=\"data row1 col16\" >1</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2127ae897c8>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(df_train['Initial'], df_train['Sex']).T.style.background_gradient(cmap = 'summer_r')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:37.839966Z",
     "start_time": "2020-06-09T06:37:37.824268Z"
    }
   },
   "outputs": [],
   "source": [
    "# 위에서 생성된 테이블을 참고하여 남자, 여자가 쓰는 Initial을 구분\n",
    "df_train['Initial'].replace(['Mlle', 'Mme', 'Ms', 'Dr', 'Major',\n",
    "                             'Lady', 'Countess', 'Jonkheer', 'Col',\n",
    "                             'Rev', 'Capt', 'Sir', 'Don', 'Dona'],\n",
    "                            ['Miss', 'Miss', 'Miss', 'Mr', 'Mr',\n",
    "                             'Mrs', 'Mrs', 'Other', 'Other',\n",
    "                             'Other', 'Mr', 'Mr', 'Mr', 'Mr'], inplace = True)\n",
    "\n",
    "df_test['Initial'].replace(['Mlle', 'Mme', 'Ms', 'Dr', 'Major',\n",
    "                            'Lady', 'Countess', 'Jonkheer', 'Col',\n",
    "                            'Rev', 'Capt', 'Sir', 'Don', 'Dona'],\n",
    "                           ['Miss', 'Miss', 'Miss', 'Mr', 'Mr',\n",
    "                            'Mrs', 'Mrs', 'Other', 'Other',\n",
    "                            'Other', 'Mr', 'Mr', 'Mr', 'Mr'], inplace = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:37.877902Z",
     "start_time": "2020-06-09T06:37:37.841961Z"
    }
   },
   "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>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "      <th>FamilySize</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Initial</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Master</th>\n",
       "      <td>414.975000</td>\n",
       "      <td>0.575000</td>\n",
       "      <td>2.625000</td>\n",
       "      <td>4.574167</td>\n",
       "      <td>2.300000</td>\n",
       "      <td>1.375000</td>\n",
       "      <td>3.340710</td>\n",
       "      <td>4.675000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Miss</th>\n",
       "      <td>411.741935</td>\n",
       "      <td>0.704301</td>\n",
       "      <td>2.284946</td>\n",
       "      <td>21.860000</td>\n",
       "      <td>0.698925</td>\n",
       "      <td>0.537634</td>\n",
       "      <td>3.123713</td>\n",
       "      <td>2.236559</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mr</th>\n",
       "      <td>455.880907</td>\n",
       "      <td>0.162571</td>\n",
       "      <td>2.381853</td>\n",
       "      <td>32.739609</td>\n",
       "      <td>0.293006</td>\n",
       "      <td>0.151229</td>\n",
       "      <td>2.651507</td>\n",
       "      <td>1.444234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mrs</th>\n",
       "      <td>456.393701</td>\n",
       "      <td>0.795276</td>\n",
       "      <td>1.984252</td>\n",
       "      <td>35.981818</td>\n",
       "      <td>0.692913</td>\n",
       "      <td>0.818898</td>\n",
       "      <td>3.443751</td>\n",
       "      <td>2.511811</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Other</th>\n",
       "      <td>564.444444</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>1.666667</td>\n",
       "      <td>45.888889</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>2.641605</td>\n",
       "      <td>1.222222</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         PassengerId  Survived    Pclass        Age     SibSp     Parch  \\\n",
       "Initial                                                                   \n",
       "Master    414.975000  0.575000  2.625000   4.574167  2.300000  1.375000   \n",
       "Miss      411.741935  0.704301  2.284946  21.860000  0.698925  0.537634   \n",
       "Mr        455.880907  0.162571  2.381853  32.739609  0.293006  0.151229   \n",
       "Mrs       456.393701  0.795276  1.984252  35.981818  0.692913  0.818898   \n",
       "Other     564.444444  0.111111  1.666667  45.888889  0.111111  0.111111   \n",
       "\n",
       "             Fare  FamilySize  \n",
       "Initial                        \n",
       "Master   3.340710    4.675000  \n",
       "Miss     3.123713    2.236559  \n",
       "Mr       2.651507    1.444234  \n",
       "Mrs      3.443751    2.511811  \n",
       "Other    2.641605    1.222222  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.groupby('Initial').mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:38.095286Z",
     "start_time": "2020-06-09T06:37:37.879860Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x2127c65c148>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_train.groupby('Initial')['Survived'].mean().plot.bar()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- 결과를 살펴보면, 여성과 관계있는 **Miss, Mrs**가 **생존률이 높은 것**을 알 수 있다."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Statistics를 사용하여 Null 값 처리**\n",
    "- 학습 데이터 셋에서 얻은 statistics를 기반으로, 테스트 데이터 셋의 Null 값을 채워줘야 한다.\n",
    "- Pandas 데이터 프레임을 다룰 때에는 boolean array를 이용해서 indexing하는 방법이 정말 편리하다!!\n",
    "    - ***```loc + boolean + column```*** 을 사용해서 값을 치환하는 방법은 자주 쓰이므로, 반드시 잘 숙지하도록 하자!!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:38.119221Z",
     "start_time": "2020-06-09T06:37:38.099274Z"
    }
   },
   "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>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "      <th>FamilySize</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Initial</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Master</th>\n",
       "      <td>414.975000</td>\n",
       "      <td>0.575000</td>\n",
       "      <td>2.625000</td>\n",
       "      <td>4.574167</td>\n",
       "      <td>2.300000</td>\n",
       "      <td>1.375000</td>\n",
       "      <td>3.340710</td>\n",
       "      <td>4.675000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Miss</th>\n",
       "      <td>411.741935</td>\n",
       "      <td>0.704301</td>\n",
       "      <td>2.284946</td>\n",
       "      <td>21.860000</td>\n",
       "      <td>0.698925</td>\n",
       "      <td>0.537634</td>\n",
       "      <td>3.123713</td>\n",
       "      <td>2.236559</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mr</th>\n",
       "      <td>455.880907</td>\n",
       "      <td>0.162571</td>\n",
       "      <td>2.381853</td>\n",
       "      <td>32.739609</td>\n",
       "      <td>0.293006</td>\n",
       "      <td>0.151229</td>\n",
       "      <td>2.651507</td>\n",
       "      <td>1.444234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mrs</th>\n",
       "      <td>456.393701</td>\n",
       "      <td>0.795276</td>\n",
       "      <td>1.984252</td>\n",
       "      <td>35.981818</td>\n",
       "      <td>0.692913</td>\n",
       "      <td>0.818898</td>\n",
       "      <td>3.443751</td>\n",
       "      <td>2.511811</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Other</th>\n",
       "      <td>564.444444</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>1.666667</td>\n",
       "      <td>45.888889</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>0.111111</td>\n",
       "      <td>2.641605</td>\n",
       "      <td>1.222222</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         PassengerId  Survived    Pclass        Age     SibSp     Parch  \\\n",
       "Initial                                                                   \n",
       "Master    414.975000  0.575000  2.625000   4.574167  2.300000  1.375000   \n",
       "Miss      411.741935  0.704301  2.284946  21.860000  0.698925  0.537634   \n",
       "Mr        455.880907  0.162571  2.381853  32.739609  0.293006  0.151229   \n",
       "Mrs       456.393701  0.795276  1.984252  35.981818  0.692913  0.818898   \n",
       "Other     564.444444  0.111111  1.666667  45.888889  0.111111  0.111111   \n",
       "\n",
       "             Fare  FamilySize  \n",
       "Initial                        \n",
       "Master   3.340710    4.675000  \n",
       "Miss     3.123713    2.236559  \n",
       "Mr       2.651507    1.444234  \n",
       "Mrs      3.443751    2.511811  \n",
       "Other    2.641605    1.222222  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Age 변수에 존재하는 Null 값들을 Age 변수의 평균값으로 대체하기 위해, 평균값들을 확인\n",
    "df_train.groupby('Initial').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:38.179071Z",
     "start_time": "2020-06-09T06:37:38.122213Z"
    }
   },
   "outputs": [],
   "source": [
    "# Null 값들을 Age 변수의 평균값으로 대체\n",
    "df_train.loc[(df_train.Age.isnull()) & (df_train.Initial == 'Mr'), 'Age'] = 33\n",
    "df_train.loc[(df_train.Age.isnull()) & (df_train.Initial == 'Mrs'), 'Age'] = 36\n",
    "df_train.loc[(df_train.Age.isnull()) & (df_train.Initial == 'Master'), 'Age'] = 5\n",
    "df_train.loc[(df_train.Age.isnull()) & (df_train.Initial == 'Miss'), 'Age'] = 22\n",
    "df_train.loc[(df_train.Age.isnull()) & (df_train.Initial == 'Other'), 'Age'] = 46\n",
    "\n",
    "df_test.loc[(df_test.Age.isnull()) & (df_test.Initial == 'Mr'), 'Age'] = 33\n",
    "df_test.loc[(df_test.Age.isnull()) & (df_test.Initial == 'Mrs'), 'Age'] = 36\n",
    "df_test.loc[(df_test.Age.isnull()) & (df_test.Initial == 'Master'), 'Age'] = 5\n",
    "df_test.loc[(df_test.Age.isnull()) & (df_test.Initial == 'Miss'), 'Age'] = 22\n",
    "df_test.loc[(df_test.Age.isnull()) & (df_test.Initial == 'Other'), 'Age'] = 46"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**2. Embarked 변수의 Null 값 처리**\n",
    "- Embarked 변수에는 Null 값이 2개 존재한다.\n",
    "- 탑승 항구 \"S\"에서 가장 많은 탑승객이 있었으므로, 간단하게 **Null 값들을 \"S\"로 대체**해주겠다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:38.190032Z",
     "start_time": "2020-06-09T06:37:38.182055Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Embarked has 2 Null values\n"
     ]
    }
   ],
   "source": [
    "print('Embarked has', sum(df_train['Embarked'].isnull()), 'Null values')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:38.201035Z",
     "start_time": "2020-06-09T06:37:38.194020Z"
    }
   },
   "outputs": [],
   "source": [
    "df_train['Embarked'].fillna('S', inplace = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 연속형 변수인 Age를 범주형 변수로 변환\n",
    "- 이처럼 연속형 변수를 범주형 변수로 변환하는 경우, **정보 손실**이 **발생**할 수 있다.\n",
    "- 그러나 본 튜토리얼에서는 다양한 방법들을 소개하는 것이 목적이므로 진행하도록 하겠다.\n",
    "- 범주형 변수로 변환해주는 방법은 여러 가지가 존재한다.\n",
    "    - 첫 번째 방법: 데이터 프레임의 indexing 방법인 ***```loc```*** 를 사용하여 직접 변환하는 방법\n",
    "    - 두 번째 방법: ***```apply```*** 를 사용해서 함수를 넣어주는 방법\n",
    "    - 참고로 두 번째 방법(***```apply```***)을 사용하는 것이 훨씬 더 간단하다!! "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:38.284777Z",
     "start_time": "2020-06-09T06:37:38.204991Z"
    }
   },
   "outputs": [],
   "source": [
    "# 1. loc 를 사용한 방법\n",
    "df_train['Age_cat'] = 0\n",
    "df_train.loc[df_train['Age'] < 10, 'Age_cat'] = 0\n",
    "df_train.loc[(10 <= df_train['Age']) & (df_train['Age'] < 20), 'Age_cat'] = 1\n",
    "df_train.loc[(20 <= df_train['Age']) & (df_train['Age'] < 30), 'Age_cat'] = 2\n",
    "df_train.loc[(30 <= df_train['Age']) & (df_train['Age'] < 40), 'Age_cat'] = 3\n",
    "df_train.loc[(40 <= df_train['Age']) & (df_train['Age'] < 50), 'Age_cat'] = 4\n",
    "df_train.loc[(50 <= df_train['Age']) & (df_train['Age'] < 60), 'Age_cat'] = 5\n",
    "df_train.loc[(60 <= df_train['Age']) & (df_train['Age'] < 70), 'Age_cat'] = 6\n",
    "df_train.loc[df_train['Age'] >= 70, 'Age_cat'] = 7\n",
    "\n",
    "df_test['Age_cat'] = 0\n",
    "df_test.loc[df_test['Age'] < 10, 'Age_cat'] = 0\n",
    "df_test.loc[(10 <= df_test['Age']) & (df_test['Age'] < 20), 'Age_cat'] = 1\n",
    "df_test.loc[(20 <= df_test['Age']) & (df_test['Age'] < 30), 'Age_cat'] = 2\n",
    "df_test.loc[(30 <= df_test['Age']) & (df_test['Age'] < 40), 'Age_cat'] = 3\n",
    "df_test.loc[(40 <= df_test['Age']) & (df_test['Age'] < 50), 'Age_cat'] = 4\n",
    "df_test.loc[(50 <= df_test['Age']) & (df_test['Age'] < 60), 'Age_cat'] = 5\n",
    "df_test.loc[(60 <= df_test['Age']) & (df_test['Age'] < 70), 'Age_cat'] = 6\n",
    "df_test.loc[df_test['Age'] >= 70, 'Age_cat'] = 7"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:38.302786Z",
     "start_time": "2020-06-09T06:37:38.287769Z"
    }
   },
   "outputs": [],
   "source": [
    "# 2. apply 를 사용한 방법 --> 훨씬 더 간단!!\n",
    "def category_age(x):\n",
    "    if x < 10:\n",
    "        return 0\n",
    "    elif x < 20:\n",
    "        return 1\n",
    "    elif x < 30:\n",
    "        return 2\n",
    "    elif x < 40:\n",
    "        return 3\n",
    "    elif x < 50:\n",
    "        return 4\n",
    "    elif x < 60:\n",
    "        return 5\n",
    "    elif x < 70:\n",
    "        return 6\n",
    "    else:\n",
    "        return 7\n",
    "    \n",
    "df_train['Age_cat2'] = df_train['Age'].apply(category_age)\n",
    "df_test['Age_cat2'] = df_test['Age'].apply(category_age)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:38.313701Z",
     "start_time": "2020-06-09T06:37:38.304723Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1번 방법, 2번 방법 둘 다 같은 결과를 내면 True 를 반환해줘야 함 ->  True\n"
     ]
    }
   ],
   "source": [
    "# 첫 번째 방법(loc)과 두 번째 방법(apply)의 결과가 동일한지 확인\n",
    "print('1번 방법, 2번 방법 둘 다 같은 결과를 내면 True 를 반환해줘야 함 -> ',\n",
    "      (df_train['Age_cat'] == df_train['Age_cat2']).all())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:38.329657Z",
     "start_time": "2020-06-09T06:37:38.316694Z"
    }
   },
   "outputs": [],
   "source": [
    "# 이제 중복되는 Age_cat2 컬럼과 원래 컬럼인 Age를 제거해주겠다\n",
    "df_train.drop(['Age', 'Age_cat2'], axis = 1, inplace = True)\n",
    "df_test.drop(['Age', 'Age_cat2'], axis = 1, inplace = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 문자열 변수인 Initial, Embarked, Sex를 수치형 변수로 변환\n",
    "- ```map()``` 메소드를 사용하면 간단하게 변환해 줄 수 있다."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Initial 변수를 수치화**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:38.344618Z",
     "start_time": "2020-06-09T06:37:38.331652Z"
    }
   },
   "outputs": [],
   "source": [
    "df_train['Initial'] = df_train['Initial'].map({'Master': 0, 'Miss': 1, 'Mr': 2, 'Mrs': 3, 'Other': 4})\n",
    "df_test['Initial'] = df_test['Initial'].map({'Master': 0, 'Miss': 1, 'Mr': 2, 'Mrs': 3, 'Other': 4})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Embarked 변수를 수치화**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:38.355588Z",
     "start_time": "2020-06-09T06:37:38.347609Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['S', 'C', 'Q'], dtype=object)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train['Embarked'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:38.368555Z",
     "start_time": "2020-06-09T06:37:38.358580Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "S    646\n",
       "C    168\n",
       "Q     77\n",
       "Name: Embarked, dtype: int64"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train['Embarked'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:38.381518Z",
     "start_time": "2020-06-09T06:37:38.371547Z"
    }
   },
   "outputs": [],
   "source": [
    "df_train['Embarked'] = df_train['Embarked'].map({'C': 0, 'Q': 1, 'S': 2})\n",
    "df_test['Embarked'] = df_test['Embarked'].map({'C': 0, 'Q': 1, 'S': 2})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:38.393486Z",
     "start_time": "2020-06-09T06:37:38.384511Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# any()를 사용하여, True가 단 하나라도 있으면 True를 반환\n",
    "# 즉, Null 값이 한 개라도 존재하면 True를 반환\n",
    "df_train['Embarked'].isnull().any()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Sex 변수를 수치화**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:38.405455Z",
     "start_time": "2020-06-09T06:37:38.395481Z"
    }
   },
   "outputs": [],
   "source": [
    "df_train['Sex'] = df_train['Sex'].map({'female': 0, 'male': 1})\n",
    "df_test['Sex'] = df_test['Sex'].map({'female': 0, 'male': 1})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**각 feature 간의 상관관계를 시각화**\n",
    "- 피어슨 상관계수가 **1 에 가까우면 양의 상관관계**, **-1 에 가까우면 음의 상관관계**이다.\n",
    "- **피어슨 상관계수가 0** 이라는 것은 상관관계가 없다는 의미가 아니라, **선형적인 상관관계가 없다는 의미**이다.\n",
    "    - 즉, **비선형적 상관관계는 존재할 수 있다!**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:39.011715Z",
     "start_time": "2020-06-09T06:37:38.408447Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x864 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "heatmap_data = df_train[['Survived', 'Pclass', 'Sex', \n",
    "                         'Fare', 'Embarked', 'FamilySize',\n",
    "                         'Initial', 'Age_cat']]\n",
    "\n",
    "colormap = plt.cm.RdBu\n",
    "plt.figure(figsize = (14, 12))\n",
    "plt.title('Pearson Correlation of Features', y = 1.05, size = 15)\n",
    "sns.heatmap(heatmap_data.astype(float).corr(), linewidths = 0.1, vmax = 1.0, square = True, \n",
    "            cmap = colormap, linecolor = 'white', annot = True, annot_kws = {'size': 16})\n",
    "\n",
    "del heatmap_data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- 앞선 EDA 과정에서 살펴봤듯이, **Sex와 Pclass 변수가 Survived 변수와 어느 정도 상관관계가 존재함**을 알 수 있다.\n",
    "- 하지만 **서로 강한 상관관계를 갖는 feature들은 존재하지 않는 것을 확인**할 수 있다.\n",
    "    - 즉, **다중 공선성을 보이는 변수들이 존재하지 않는다**는 의미이다.\n",
    "    - 다시 말해, 모델 생성에 있어서 **불필요한 feature가 없다**는 말이다."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### One-Hot Encoding(원-핫 인코딩)\n",
    "- 수치화를 시켜준 범주형 데이터를 그대로 모델에 넣어줘도 되지만, 모델의 성능을 높여주기 위해서 **one-hot encoding** 작업을 수행해주겠다.\n",
    "    - 원-핫 인코딩이라는게 쉽게 말하면 **가변수(더미변수)를 만들어주겠다**는 말이랑 동일하다.\n",
    "    - **Pandas**의 ```get_dummies()```를 사용하면 쉽게 수행해 줄 수 있다.\n",
    "        - ```prefix``` 옵션을 사용하면, **가변수에 공통으로 접두사를 추가**할 수 있다.\n",
    "        - ```drop_first = True``` 옵션을 설정하면, **가변수의 첫 번째 변수를 자동으로 삭제**해준다.\n",
    "            - 즉, **가변수 함정(dummy_trap)을 피할 수 있게 해준다 --> 총 (k-1)개의 가변수 생성!!**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Initial 변수에 대해서 One-Hot Encoding 처리**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:39.036646Z",
     "start_time": "2020-06-09T06:37:39.015703Z"
    }
   },
   "outputs": [],
   "source": [
    "df_train = pd.get_dummies(df_train, columns = ['Initial'], prefix = 'Initial')\n",
    "df_test = pd.get_dummies(df_test, columns = ['Initial'], prefix = 'Initial')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:39.066568Z",
     "start_time": "2020-06-09T06:37:39.040636Z"
    }
   },
   "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>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "      <th>FamilySize</th>\n",
       "      <th>Age_cat</th>\n",
       "      <th>Initial_0</th>\n",
       "      <th>Initial_1</th>\n",
       "      <th>Initial_2</th>\n",
       "      <th>Initial_3</th>\n",
       "      <th>Initial_4</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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>1.981001</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</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>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>4.266662</td>\n",
       "      <td>C85</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</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>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>2.070022</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</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>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>3.972177</td>\n",
       "      <td>C123</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</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>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>373450</td>\n",
       "      <td>2.085672</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</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  SibSp  Parch  \\\n",
       "0                            Braund, Mr. Owen Harris    1      1      0   \n",
       "1  Cumings, Mrs. John Bradley (Florence Briggs Th...    0      1      0   \n",
       "2                             Heikkinen, Miss. Laina    0      0      0   \n",
       "3       Futrelle, Mrs. Jacques Heath (Lily May Peel)    0      1      0   \n",
       "4                           Allen, Mr. William Henry    1      0      0   \n",
       "\n",
       "             Ticket      Fare Cabin  Embarked  FamilySize  Age_cat  Initial_0  \\\n",
       "0         A/5 21171  1.981001   NaN         2           2        2          0   \n",
       "1          PC 17599  4.266662   C85         0           2        3          0   \n",
       "2  STON/O2. 3101282  2.070022   NaN         2           1        2          0   \n",
       "3            113803  3.972177  C123         2           2        3          0   \n",
       "4            373450  2.085672   NaN         2           1        3          0   \n",
       "\n",
       "   Initial_1  Initial_2  Initial_3  Initial_4  \n",
       "0          0          1          0          0  \n",
       "1          0          0          1          0  \n",
       "2          1          0          0          0  \n",
       "3          0          0          1          0  \n",
       "4          0          1          0          0  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Embarked 변수에 대해서 One-Hot Encoding 처리**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:39.093495Z",
     "start_time": "2020-06-09T06:37:39.069559Z"
    }
   },
   "outputs": [],
   "source": [
    "df_train = pd.get_dummies(df_train, columns = ['Embarked'], prefix = 'Embarked')\n",
    "df_test = pd.get_dummies(df_test, columns = ['Embarked'], prefix = 'Embarked')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:39.123415Z",
     "start_time": "2020-06-09T06:37:39.096486Z"
    }
   },
   "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>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>FamilySize</th>\n",
       "      <th>Age_cat</th>\n",
       "      <th>Initial_0</th>\n",
       "      <th>Initial_1</th>\n",
       "      <th>Initial_2</th>\n",
       "      <th>Initial_3</th>\n",
       "      <th>Initial_4</th>\n",
       "      <th>Embarked_0</th>\n",
       "      <th>Embarked_1</th>\n",
       "      <th>Embarked_2</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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>1.981001</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</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>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>4.266662</td>\n",
       "      <td>C85</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</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>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>2.070022</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</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>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>3.972177</td>\n",
       "      <td>C123</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</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>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>373450</td>\n",
       "      <td>2.085672</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</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  SibSp  Parch  \\\n",
       "0                            Braund, Mr. Owen Harris    1      1      0   \n",
       "1  Cumings, Mrs. John Bradley (Florence Briggs Th...    0      1      0   \n",
       "2                             Heikkinen, Miss. Laina    0      0      0   \n",
       "3       Futrelle, Mrs. Jacques Heath (Lily May Peel)    0      1      0   \n",
       "4                           Allen, Mr. William Henry    1      0      0   \n",
       "\n",
       "             Ticket      Fare Cabin  FamilySize  Age_cat  Initial_0  \\\n",
       "0         A/5 21171  1.981001   NaN           2        2          0   \n",
       "1          PC 17599  4.266662   C85           2        3          0   \n",
       "2  STON/O2. 3101282  2.070022   NaN           1        2          0   \n",
       "3            113803  3.972177  C123           2        3          0   \n",
       "4            373450  2.085672   NaN           1        3          0   \n",
       "\n",
       "   Initial_1  Initial_2  Initial_3  Initial_4  Embarked_0  Embarked_1  \\\n",
       "0          0          1          0          0           0           0   \n",
       "1          0          0          1          0           1           0   \n",
       "2          1          0          0          0           0           0   \n",
       "3          0          0          1          0           0           0   \n",
       "4          0          1          0          0           0           0   \n",
       "\n",
       "   Embarked_2  \n",
       "0           1  \n",
       "1           0  \n",
       "2           1  \n",
       "3           1  \n",
       "4           1  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- 이제 **pandas**의 ```get_dummies()```를 사용하여 손쉽게 **One-Hot Encoding**을 수행해주었다.\n",
    "- 추가적으로 **```Labelencoder``` + ```OneHotencoder```**를 이용해도 One-Hot Encoding이 가능하다.\n",
    "- 그러나 본 튜토리얼에서는 그냥 **pandas**의 ```get_dummies()```를 사용하겠다."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 불필요한 컬럼 제거"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:39.139372Z",
     "start_time": "2020-06-09T06:37:39.127404Z"
    }
   },
   "outputs": [],
   "source": [
    "df_train.drop(['PassengerId', 'Name', 'SibSp', 'Parch', 'Ticket', 'Cabin'], axis = 1, inplace = True)\n",
    "df_test.drop(['PassengerId', 'Name', 'SibSp', 'Parch', 'Ticket', 'Cabin'], axis = 1, inplace = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:39.164305Z",
     "start_time": "2020-06-09T06:37:39.141366Z"
    }
   },
   "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>Sex</th>\n",
       "      <th>Fare</th>\n",
       "      <th>FamilySize</th>\n",
       "      <th>Age_cat</th>\n",
       "      <th>Initial_0</th>\n",
       "      <th>Initial_1</th>\n",
       "      <th>Initial_2</th>\n",
       "      <th>Initial_3</th>\n",
       "      <th>Initial_4</th>\n",
       "      <th>Embarked_0</th>\n",
       "      <th>Embarked_1</th>\n",
       "      <th>Embarked_2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1.981001</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4.266662</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2.070022</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3.972177</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2.085672</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Survived  Pclass  Sex      Fare  FamilySize  Age_cat  Initial_0  Initial_1  \\\n",
       "0         0       3    1  1.981001           2        2          0          0   \n",
       "1         1       1    0  4.266662           2        3          0          0   \n",
       "2         1       3    0  2.070022           1        2          0          1   \n",
       "3         1       1    0  3.972177           2        3          0          0   \n",
       "4         0       3    1  2.085672           1        3          0          0   \n",
       "\n",
       "   Initial_2  Initial_3  Initial_4  Embarked_0  Embarked_1  Embarked_2  \n",
       "0          1          0          0           0           0           1  \n",
       "1          0          1          0           1           0           0  \n",
       "2          0          0          0           0           0           1  \n",
       "3          0          1          0           0           0           1  \n",
       "4          1          0          0           0           0           1  "
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:39.189238Z",
     "start_time": "2020-06-09T06:37:39.166299Z"
    }
   },
   "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>Sex</th>\n",
       "      <th>Fare</th>\n",
       "      <th>FamilySize</th>\n",
       "      <th>Age_cat</th>\n",
       "      <th>Initial_0</th>\n",
       "      <th>Initial_1</th>\n",
       "      <th>Initial_2</th>\n",
       "      <th>Initial_3</th>\n",
       "      <th>Initial_4</th>\n",
       "      <th>Embarked_0</th>\n",
       "      <th>Embarked_1</th>\n",
       "      <th>Embarked_2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2.057860</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1.945910</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2.270836</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2.159003</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2.508582</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Pclass  Sex      Fare  FamilySize  Age_cat  Initial_0  Initial_1  \\\n",
       "0       3    1  2.057860           1        3          0          0   \n",
       "1       3    0  1.945910           2        4          0          0   \n",
       "2       2    1  2.270836           1        6          0          0   \n",
       "3       3    1  2.159003           1        2          0          0   \n",
       "4       3    0  2.508582           3        2          0          0   \n",
       "\n",
       "   Initial_2  Initial_3  Initial_4  Embarked_0  Embarked_1  Embarked_2  \n",
       "0          1          0          0           0           1           0  \n",
       "1          0          1          0           0           0           1  \n",
       "2          1          0          0           0           1           0  \n",
       "3          1          0          0           0           0           1  \n",
       "4          0          1          0           0           0           1  "
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 머신러닝 모델 생성 및 예측\n",
    "- 본 튜토리얼은 **0 과 1 로 이루어진 target 변수에 대한 예측**을 수행하는 모델을 만드는 것이다.\n",
    "    - 즉, **이진 분류 문제**라고 할 수 있다.\n",
    "- 우선 **학습 데이터 셋**에서 **Survived를 제외한 input**을 가지고, **모델 최적화**를 수행해 줄 것이다.\n",
    "- 그 후에 모델이 학습하지 않았던 **테스트 데이터 셋을 input**으로 주어서, **테스트 데이터 셋의 각 탑승객의 생존 여부를 예측**해보겠다.\n",
    "- 본 튜토리얼에서는 **랜덤 포레스트** 모델을 사용할 것이다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:39.460753Z",
     "start_time": "2020-06-09T06:37:39.193229Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn import metrics\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 데이터 셋 분할\n",
    "1. **모델 학습**에 사용할 ***train dataset***\n",
    "2. **모델 평가**에 사용할 ***valid dataset***\n",
    "    - 좋은 모델을 만들기 위해서는 ***valid dataset*** 을 따로 만들어서 모델을 평가해주어야 한다.\n",
    "    - 마치 축구대표팀이 **팀훈련(train)**을 하고 바로 **월드컵(test)**에 나가는 것이 아니라, **팀훈련(train)**을 한 다음 **평가전(valid)**을 거쳐 **팀의 훈련 정도(학습 정도)를 확인**하고 **월드컵(test)**에 나가는 것과 비슷하다.\n",
    "3. **모델 예측**에 사용할 ***test dataset***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- 먼저 **학습 데이터 셋(train dataset)**과 **target label(Survived)**을 **분리**하겠다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:39.469733Z",
     "start_time": "2020-06-09T06:37:39.462752Z"
    }
   },
   "outputs": [],
   "source": [
    "X_train = df_train.drop('Survived', axis = 1).values\n",
    "target_label = df_train['Survived'].values\n",
    "X_test = df_test.values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- ```train_test_split()```을 사용하여 데이터 셋을 쉽게 분할할 수 있다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:39.478709Z",
     "start_time": "2020-06-09T06:37:39.471729Z"
    }
   },
   "outputs": [],
   "source": [
    "X_tr, X_vld, y_tr, y_vld = train_test_split(X_train, target_label, test_size = 0.3, random_state = 2020)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 모델 학습 및 예측\n",
    "- 먼저 모델 객체를 만들고, ```fit()``` 메소드로 학습시키겠다.\n",
    "- 그 다음, **valid dataset**을 input으로 넣어줘서 예측 값을 구해보도록 하겠다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:39.677180Z",
     "start_time": "2020-06-09T06:37:39.481702Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "총 268명 중, 81.34% 정확도로 생존을 맞춤\n"
     ]
    }
   ],
   "source": [
    "model = RandomForestClassifier()\n",
    "model.fit(X_tr, y_tr)\n",
    "prediction = model.predict(X_vld)\n",
    "\n",
    "print('총 {}명 중, {:.2f}% 정확도로 생존을 맞춤'.format(y_vld.shape[0], 100 * metrics.accuracy_score(prediction, y_vld)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- 파라미터 튜닝을 하지도 않았는데 **약 82 %의 정확도**가 나왔다!!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature 중요도 확인\n",
    "- 생성된 예측 모델이 어떤 feature의 영향을 많이 받았는지 확인해보는 작업을 수행하겠다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:39.695166Z",
     "start_time": "2020-06-09T06:37:39.679173Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "from pandas import Series\n",
    "\n",
    "feature_importance = model.feature_importances_\n",
    "Series_feat_imp = Series(feature_importance, index = df_test.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:39.704109Z",
     "start_time": "2020-06-09T06:37:39.697125Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Pclass', 'Sex', 'Fare', 'FamilySize', 'Age_cat', 'Initial_0',\n",
       "       'Initial_1', 'Initial_2', 'Initial_3', 'Initial_4', 'Embarked_0',\n",
       "       'Embarked_1', 'Embarked_2'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:40.007331Z",
     "start_time": "2020-06-09T06:37:39.706103Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (8, 8))\n",
    "Series_feat_imp.sort_values(ascending = True).plot.barh()\n",
    "plt.xlabel('Feature importance')\n",
    "plt.ylabel('Feature')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- 결과를 보면, 가장 중요도가 높은 feature는 **Fare**이다.\n",
    "- 그 다음으로 **Sex, Initial_2, Age_cat**이 뒤를 이었다.\n",
    "- 만약 좀 더 정확도가 높은 모델을 얻고 싶다면, **feature selection**을 수행해주면 된다."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 테스트 데이터 셋에 대해서 예측을 수행\n",
    "- 이제 모델이 학습하지 않았던 테스트 데이터 셋을 input으로 넣어주고, 각 탑승객의 생존 여부를 예측해보겠다.\n",
    "- Kaggle에서 준 파일인 **gender_submission.csv** 파일을 읽어와서 제출할 준비를 하겠다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:40.021259Z",
     "start_time": "2020-06-09T06:37:40.009291Z"
    }
   },
   "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>1</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>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived\n",
       "0          892         0\n",
       "1          893         1\n",
       "2          894         0\n",
       "3          895         0\n",
       "4          896         1"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "submission = pd.read_csv('../titanic/gender_submission.csv')\n",
    "submission.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:40.046193Z",
     "start_time": "2020-06-09T06:37:40.024255Z"
    }
   },
   "outputs": [],
   "source": [
    "prediction = model.predict(X_test)\n",
    "submission['Survived'] = prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-06-09T06:37:40.061153Z",
     "start_time": "2020-06-09T06:37:40.050199Z"
    }
   },
   "outputs": [],
   "source": [
    "submission.to_csv('./my_first_submission.csv', index = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**여기까지 해서 타이타닉 튜토리얼을 마무리하도록 하겠다!**\n",
    "\n",
    "**비록 본 튜토리얼에서는 기본적인 데이터 전처리 과정과 단순한 모델을 사용해서 예측 모델을 만들었지만, 앞으로 공부를 이어나가면서 더욱 참신한 아이디어로 머신러닝 모델을 만들어 볼 예정이다.**"
   ]
  }
 ],
 "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": "3.7.6"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}