02_PreProcessing/main02.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 前處理-資料清理\n",
    "1. missing data\n",
    "    - 平均值、中位數、屆在平均值與標準差之間的亂數\n",
    "    \n",
    "2. range差異過大的資料\n",
    "    - 偵測(identity): PCA\n",
    "    \n",
    "    - 處理: 取log e.g. 10^1 => 1, 10^6 => 6\n",
    "    \n",
    "4. 資料不一致的問題\n",
    "    - domain knowledge e.g. 年紀為負的\n",
    "    \n",
    "5. 正規化(Normalize)\n",
    "    - L1 Norm(穩定: 水平調整較少): \n",
    "        - $\\|𝑥\\|_{1𝑖} = \\dfrac{𝑥_𝑖}{\\sum{|𝑥_𝑖|}}$\n",
    "    - L2 Norm(強健: 較能對抗outlier): \n",
    "        - $\\|𝑥\\|_{2𝑖}=\\dfrac{𝑥_𝑖}{\\sqrt{(\\sum{𝑥_𝑖^2)}}}$\n",
    "    - Norm to 0 ~ 1\n",
    "        - $\\|𝑥\\|=\\dfrac{𝑥_𝑖 - min(x)}{ max(x) - min(x)}$\n",
    "\n",
    "6. 類別型資料的處理\n",
    "    - 自然語言(NLP): 先轉成類別型資料\n",
    "    - onehot encoding\n",
    "    \n",
    "| 0 | 1 | 2 |\n",
    "|---|---|---|\n",
    "|[1,0,0]|[0,1,0]|[0,0,1]|\n",
    "\n",
    "7. Feature 產生工具\n",
    "    - PolynomialFeatures: (𝑋_1, 𝑋_2)=> (1, 𝑋_1, 𝑋_2, 𝑋_1^2, 𝑋_1 𝑋_2, 𝑋_2^2)\n",
    "    \n",
    "8. 議題:\n",
    "    - 請問取log與normalize有什麼差別?\n",
    "    \n",
    "    \n",
    "    \n",
    "    \n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# IMPORT & DATA\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os             # for directory & file operation\n",
    "import re             # for text detection\n",
    "import pandas as pd   # excel in python\n",
    "import numpy as np    # matrix operation\n",
    "from collections import Counter     # count num for each items in the list\n",
    "from sklearn import preprocessing   # Normalize, OneHotEncoding, PolynomialFeatures\n",
    "\n",
    "# PCA calculation & Drawing\n",
    "import matplotlib.pyplot as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "from sklearn.decomposition import PCA\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## os \n",
    "```\n",
    "'\\n' => 換行\n",
    "'\\t' => tab\n",
    "'\\\\' => \\\n",
    "```\n",
    "\n",
    "```\n",
    "# os.listdir # list files\n",
    "# os.path.join # join file paths\n",
    "# os.path.isdir \n",
    "# os.path.isfile \n",
    "# os.remove\n",
    "# os.mkdir => create new directory\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### pandas indexing\n",
    "\n",
    "```\n",
    "type(df) => DataFrame\n",
    "df.columns => call columns\n",
    "df['Survived'] => indexing all rows for a specific column\n",
    "type(df['Survived']) => Series\n",
    "df[['PassengerId', 'Survived']] => indexing all rows for specific columns\n",
    "df.loc[10] => indexing all column for a specific row\n",
    "df.loc[10:12] => indexing all column for specific rows\n",
    "df.loc[10:12, ['PassengerId', 'Survived']] => indexing both columns and rows\n",
    "df.loc[df['Survived'] == 1] => indexing by condition\n",
    "df[(df['Survived'] == 1) & (df['Age'] > 30)] => indexing by multiple conditions\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Ways to run for loop (DataFrame)\n",
    "```\n",
    "# version 0\n",
    "indexer = []\n",
    "for idx, row in df.iterrows():\n",
    "    if 'Mr' in row['Name']:\n",
    "        indexer.append(True)\n",
    "    else:\n",
    "        indexer.append(False)\n",
    "df[indexer]\n",
    "\n",
    "# version 1\n",
    "def apply_fun(row):\n",
    "    if 'Mr' in row['Name']:\n",
    "        return True\n",
    "    else:\n",
    "        return False\n",
    "\n",
    "# version 2\n",
    "def apply_fun(row):\n",
    "    return 'Mr' in row['Name']\n",
    "\n",
    "# version 3\n",
    "apply_fun = lambda row:'Mr' in row['Name']\n",
    "df[df.apply(apply_fun, axis=1)]\n",
    "\n",
    "# version 4\n",
    "df[df.apply(lambda row:'Mr' in row['Name'], axis=1)]\n",
    "```\n",
    "\n",
    "### Ways to run for loop (Series)\n",
    "```\n",
    "# def cal_len(name):\n",
    "#     return len(name)\n",
    "# df['Name'].apply(cal_len) #把function apply 到每一個element上\n",
    "# df['Name'].apply(lambda name:len(name)) #把function apply 到每一個element上\n",
    "df['Name'].apply(lambda x:len(x)) #把function apply 到每一個element上\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 891 entries, 0 to 890\n",
      "Data columns (total 12 columns):\n",
      "PassengerId    891 non-null int64\n",
      "Survived       891 non-null int64\n",
      "Pclass         891 non-null int64\n",
      "Name           891 non-null object\n",
      "Sex            891 non-null object\n",
      "Age            714 non-null float64\n",
      "SibSp          891 non-null int64\n",
      "Parch          891 non-null int64\n",
      "Ticket         891 non-null object\n",
      "Fare           891 non-null float64\n",
      "Cabin          204 non-null object\n",
      "Embarked       889 non-null object\n",
      "dtypes: float64(2), int64(5), object(5)\n",
      "memory usage: 83.7+ KB\n"
     ]
    }
   ],
   "source": [
    "df = pd.read_csv('train.csv')\n",
    "\n",
    "# 請查看df.info()\n",
    "# 並找出共有幾種型別，以及哪一些欄位有null值\n",
    "#=============your works starts===============#\n",
    "df_info = df.info()\n",
    "#==============your works ends================#"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>count</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>714.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>mean</td>\n",
       "      <td>446.000000</td>\n",
       "      <td>0.383838</td>\n",
       "      <td>2.308642</td>\n",
       "      <td>29.699118</td>\n",
       "      <td>0.523008</td>\n",
       "      <td>0.381594</td>\n",
       "      <td>32.204208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>std</td>\n",
       "      <td>257.353842</td>\n",
       "      <td>0.486592</td>\n",
       "      <td>0.836071</td>\n",
       "      <td>14.526497</td>\n",
       "      <td>1.102743</td>\n",
       "      <td>0.806057</td>\n",
       "      <td>49.693429</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>min</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.420000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>25%</td>\n",
       "      <td>223.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>20.125000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7.910400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>50%</td>\n",
       "      <td>446.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>14.454200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>75%</td>\n",
       "      <td>668.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>38.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>31.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>max</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>512.329200</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       PassengerId    Survived      Pclass         Age       SibSp  \\\n",
       "count   891.000000  891.000000  891.000000  714.000000  891.000000   \n",
       "mean    446.000000    0.383838    2.308642   29.699118    0.523008   \n",
       "std     257.353842    0.486592    0.836071   14.526497    1.102743   \n",
       "min       1.000000    0.000000    1.000000    0.420000    0.000000   \n",
       "25%     223.500000    0.000000    2.000000   20.125000    0.000000   \n",
       "50%     446.000000    0.000000    3.000000   28.000000    0.000000   \n",
       "75%     668.500000    1.000000    3.000000   38.000000    1.000000   \n",
       "max     891.000000    1.000000    3.000000   80.000000    8.000000   \n",
       "\n",
       "            Parch        Fare  \n",
       "count  891.000000  891.000000  \n",
       "mean     0.381594   32.204208  \n",
       "std      0.806057   49.693429  \n",
       "min      0.000000    0.000000  \n",
       "25%      0.000000    7.910400  \n",
       "50%      0.000000   14.454200  \n",
       "75%      0.000000   31.000000  \n",
       "max      6.000000  512.329200  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 請查看df.describe()\n",
    "# 請透過mean關注每一個變數的scale\n",
    "#=============your works starts===============#\n",
    "df_describe = df.describe()\n",
    "#==============your works ends================#\n",
    "df_describe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>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",
       "      <td>0</td>\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",
       "      <td>1</td>\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",
       "      <td>2</td>\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",
       "      <td>3</td>\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",
       "      <td>4</td>\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": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#請透過head()查看df的頭5行\n",
    "#=============your works starts===============#\n",
    "df_head = df.head()\n",
    "#==============your works ends================#\n",
    "\n",
    "df_head"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# PREPROCESSING"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Age - Fill in missing values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## np.Nan & None Detection\n",
    "```\n",
    "df['Age'][888] == None => False\n",
    "df['Age'][888] is None => False\n",
    "df['Age'][888] == np.nan => False\n",
    "df['Age'][888] is np.nan => False\n",
    "None == None => True\n",
    "None is None => True\n",
    "np.nan == np.nan => False\n",
    "np.nan is np.nan => True\n",
    "pd.isnull(np.nan) => True\n",
    "pd.isnull(None) => True\n",
    "pd.notnull(np.nan) => False\n",
    "pd.notnull(None) => False\n",
    "np.isnan(np.nan) => True\n",
    "np.isnan(None) => ERROR\n",
    "```\n",
    "\n",
    "# conclusion\n",
    "    - np.isnan or pd.isnull can be used all the time\n",
    "    - == None cannot detect np.nan\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Student\\Projects\\IIIMaterial\\venv\\lib\\site-packages\\ipykernel_launcher.py:3: FutureWarning: `Series.plot()` should not be called with positional arguments, only keyword arguments. The order of positional arguments will change in the future. Use `Series.plot(kind='hist')` instead of `Series.plot('hist',)`.\n",
      "  This is separate from the ipykernel package so we can avoid doing imports until\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 查看Age的分布狀況(hint: plot(), kind='hist')\n",
    "#=============your works starts===============#\n",
    "age_ax = df['Age'].plot('hist')\n",
    "#==============your works ends================#\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## mean calculation\n",
    "```\n",
    "df['Age'].mean() # pandas default mean\n",
    "np.mean(df['Age']) # numpy mean apply on series (exclude nan)\n",
    "np.mean(df['Age'].values) # numpy mean apply on numpy array (include nan)\n",
    "np.nanmean(df['Age'].values) # numpy mean apply on numpy array (exclude nan)\n",
    "np.mean(df.loc[pd.notnull(df['Age']), 'Age']) # numpy mean apply on series exclude nan explicitly\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "avg_age 29.69911764705882\n"
     ]
    }
   ],
   "source": [
    "# 作法一: 取平均值\n",
    "#=============your works starts===============#\n",
    "avg_age = df['Age'].mean() \n",
    "#==============your works ends================#\n",
    "\n",
    "\n",
    "print(\"avg_age\", avg_age) \n",
    "# avg_age 29.69911764705882"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "median_age 28.0\n"
     ]
    }
   ],
   "source": [
    "# 作法二: 取中位數\n",
    "#=============your works starts===============#\n",
    "median_age = df['Age'].median()\n",
    "#==============your works ends================#\n",
    "\n",
    "\n",
    "print(\"median_age\", median_age) \n",
    "# median_age 28.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "random_age [23 41 37 17 31 20 28 24 16 42 33 19 22 20 29 15 32 16 35 40 35 34 26 27\n",
      " 37 28 30 23 31 33 42 30 25 21 29 15 21 16 39 39 21 31 31 37 31 30 23 41\n",
      " 30 35 33 21 31 28 39 37 31 29 29 40 16 43 20 29 36 22 27 41 32 24 35 23\n",
      " 29 43 33 43 31 34 34 28 27 40 29 35 27 20 40 37 16 29 29 39 20 17 20 35\n",
      " 24 42 34 33 26 38 42 31 30 40 34 16 35 16 34 24 43 29 22 29 20 43 29 38\n",
      " 37 39 35 42 40 19 32 17 25 36 15 26 31 23 19 24 34 39 39 19 17 28 16 35\n",
      " 20 16 29 18 34 43 16 28 30 42 27 25 36 19 22 43 37 38 30 15 32 38 41 21\n",
      " 26 33 20 19 21 29 40 30 28]\n",
      "len(random_age) 177\n"
     ]
    }
   ],
   "source": [
    "# 作法三: 用相同的分布產生亂數塞入(hint:使用np.random.randint, min: mean - std, max: mean + std)\n",
    "np.random.seed(1212)\n",
    "#=============your works starts===============#\n",
    "std = df['Age'].std()\n",
    "mean = df['Age'].mean()\n",
    "size = sum(pd.isnull(df['Age']))\n",
    "random_age = np.random.randint(mean - std, mean + std, size=size)\n",
    "#==============your works ends================#\n",
    "\n",
    "\n",
    "print(\"random_age\", random_age) \n",
    "print(\"len(random_age)\", len(random_age)) \n",
    "# random_age [23 41 37 17 31 20 28 24 16 42 33 19 22 20 29 15 32 16 35 40 35 34 26 27\n",
    "# len(random_age) 177"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": true
   },
   "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>avg_age</th>\n",
       "      <th>median_age</th>\n",
       "      <th>random_age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>29.699118</td>\n",
       "      <td>28.0</td>\n",
       "      <td>23.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17</td>\n",
       "      <td>29.699118</td>\n",
       "      <td>28.0</td>\n",
       "      <td>41.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>19</td>\n",
       "      <td>29.699118</td>\n",
       "      <td>28.0</td>\n",
       "      <td>37.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>26</td>\n",
       "      <td>29.699118</td>\n",
       "      <td>28.0</td>\n",
       "      <td>17.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>28</td>\n",
       "      <td>29.699118</td>\n",
       "      <td>28.0</td>\n",
       "      <td>31.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      avg_age  median_age  random_age\n",
       "5   29.699118        28.0        23.0\n",
       "17  29.699118        28.0        41.0\n",
       "19  29.699118        28.0        37.0\n",
       "26  29.699118        28.0        17.0\n",
       "28  29.699118        28.0        31.0"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['avg_age'] = df['Age']\n",
    "df.loc[pd.isnull(df['Age']), 'avg_age'] = avg_age\n",
    "df['median_age'] = df['Age']\n",
    "df.loc[pd.isnull(df['Age']), 'median_age'] = median_age\n",
    "df['random_age'] = df['Age']\n",
    "df.loc[pd.isnull(df['Age']), 'random_age'] = random_age\n",
    "\n",
    "df.loc[pd.isnull(df['Age']), ['avg_age', 'median_age', 'random_age']].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Age - Normalize\n",
    "- L1 Normalization: \n",
    "    - $|| x ||_1 = \\sum_i |x_i| = |x_1| + |x_2| + \\ldots + |x_i|$\n",
    "    - $\\|𝑥\\|_{1𝑖} = \\dfrac{𝑥_𝑖}{\\sum{|𝑥_𝑖|}}$\n",
    "- L2 Normalization: \n",
    "    - $|| x ||_2 = \\sqrt{\\left(\\sum_i x_i^2\\right)} = \\sqrt{x_1^2 + x_2^2 + \\ldots + x_i^2}$\n",
    "    - $\\|𝑥\\|_{2𝑖}=\\dfrac{𝑥_𝑖}{\\sqrt{(\\sum{𝑥_𝑖^2)}}}$\n",
    "- Normalization to 0~1: \n",
    "    - $\\|𝑥\\|_{01i}=\\dfrac{𝑥_𝑖 - \\min(x)}{\\max(x) - \\min(x)}$\n",
    "\n",
    "- difference\n",
    "![l1_l2_norm](./imgs/l1_l2_norm.png)\n",
    "![normalize_dif](./imgs/normalize_dif.jpg)\n",
    "\n",
    "- quote\n",
    "> L1是蓝色的线，L2是红色的线，很明显，L1的分布对极端值更能容忍。那么如果数据损失项使用L1 Norm，很明显，L1 Norm对outlier没有L2 Norm那么敏感；如果正则化损失项使用L1的话，那么使学习到的参数倾向于稀疏，使用L2 Norm则没有这种倾向。\n",
    "> \n",
    "> 实践中，根据Quaro的data scientist Xavier Amatriain 的经验，实际应用过程中，L1 nrom几乎没有比L2 norm表现好的时候，优先使用L2 norm是比较好的选择。\n",
    "> \n",
    "> \n",
    "> 作者：知乎用户\n",
    "> 链接：https://www.zhihu.com/question/26485586/answer/89215997\n",
    "> 来源：知乎\n",
    "> 著作权归作者所有。商业转载请联系作者获得授权，非商业转载请注明出处。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.125 -0.125  0.25 ]\n",
      " [ 0.25   0.     0.   ]\n",
      " [ 0.     0.125 -0.125]]\n"
     ]
    }
   ],
   "source": [
    "# 請寫出L1 Normaliaze的function(np.abs, np.sum)\n",
    "def normalize_l1(X):\n",
    "    \"\"\"if type(X) == np.array, and X has two dimensiions\"\"\"\n",
    "    #=============your works starts===============#\n",
    "    l1_x = np.sum(np.abs(X))\n",
    "    X = X / l1_x\n",
    "    #==============your works ends================#\n",
    "    return X\n",
    "\n",
    "X = [[ 1., -1.,  2.],\n",
    "     [ 2.,  0.,  0.],\n",
    "     [ 0.,  1., -1.]]\n",
    "X_norm_l1 = normalize_l1(X)\n",
    "print(X_norm_l1)\n",
    "# [[ 0.125 -0.125  0.25 ]\n",
    "#  [ 0.25   0.     0.   ]\n",
    "#  [ 0.     0.125 -0.125]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.28867513 -0.28867513  0.57735027]\n",
      " [ 0.57735027  0.          0.        ]\n",
      " [ 0.          0.28867513 -0.28867513]]\n"
     ]
    }
   ],
   "source": [
    "# 請寫出L2 Normaliaze的function(np.square, np.sum, np.sqrt)\n",
    "def normalize_l2(X):\n",
    "    \"\"\"if type(X) == np.array, and X has two dimensiions\"\"\"\n",
    "    #=============your works starts===============#\n",
    "    l2_x = np.sqrt(np.sum(np.square(X)))\n",
    "    X = X / l2_x\n",
    "    #==============your works ends================#\n",
    "    return X\n",
    "\n",
    "X_norm_l2 = normalize_l2(X)\n",
    "print(X_norm_l2)\n",
    "# [[ 0.28867513 -0.28867513  0.57735027]\n",
    "#  [ 0.57735027  0.          0.        ]\n",
    "#  [ 0.          0.28867513 -0.28867513]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.66666667 0.         1.        ]\n",
      " [1.         0.33333333 0.33333333]\n",
      " [0.33333333 0.66666667 0.        ]]\n"
     ]
    }
   ],
   "source": [
    "# 請寫出normalize_to_01的function(np.max, np.min)\n",
    "def normalize_to_01(X):\n",
    "    \"\"\"if type(X) == np.array, and X has two dimensiions\"\"\"\n",
    "    #=============your works starts===============#\n",
    "    min_x = np.min(X)\n",
    "    max_x = np.max(X)\n",
    "    X = (X - min_x) / (max_x - min_x)\n",
    "    #==============your works ends================#\n",
    "    return X\n",
    "\n",
    "X_norm_01 = normalize_to_01(X)\n",
    "print(X_norm_01)\n",
    "# [[ 0.66666667  0.          1.        ]\n",
    "#  [ 1.          0.33333333  0.33333333]\n",
    "#  [ 0.33333333  0.66666667  0.        ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.125 -0.125  0.25 ]\n",
      " [ 0.25   0.     0.   ]\n",
      " [ 0.     0.125 -0.125]]\n",
      "[[ 0.25 -0.25  0.5 ]\n",
      " [ 1.    0.    0.  ]\n",
      " [ 0.    0.5  -0.5 ]]\n"
     ]
    }
   ],
   "source": [
    "X = [[ 1., -1.,  2.],\n",
    "     [ 2.,  0.,  0.],\n",
    "     [ 0.,  1., -1.]]\n",
    "\n",
    "# 請使用preprocessing.normalize(X, norm='l1')比較，與我們自己寫的normalize function有什麼不同\n",
    "#=============your works starts===============#\n",
    "X_normalized_self = normalize_l1(X)\n",
    "X_normalized_sklearn = preprocessing.normalize(X, norm='l1')\n",
    "#==============your works ends================#\n",
    "\n",
    "print(X_normalized_self)\n",
    "print(X_normalized_sklearn)\n",
    "# sklearn l1_norm\n",
    "# [[ 0.25 -0.25  0.5 ]\n",
    "#  [ 1.    0.    0.  ]\n",
    "#  [ 0.    0.5  -0.5 ]]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1.0, -1.0, 2.0], [2.0, 0.0, 0.0], [0.0, 1.0, -1.0]]\n",
      "[[ 0.125 -0.125  0.25 ]\n",
      " [ 0.25   0.     0.   ]\n",
      " [ 0.     0.125 -0.125]]\n",
      "[[ 0.28867513 -0.28867513  0.57735027]\n",
      " [ 0.57735027  0.          0.        ]\n",
      " [ 0.          0.28867513 -0.28867513]]\n",
      "[[0.66666667 0.         1.        ]\n",
      " [1.         0.33333333 0.33333333]\n",
      " [0.33333333 0.66666667 0.        ]]\n"
     ]
    }
   ],
   "source": [
    "print(X[:10])\n",
    "print(X_norm_l1[:10])\n",
    "print(X_norm_l2[:10])\n",
    "print(X_norm_01[:10])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# X = np.random.normal(size=100000)\n",
    "# X_norm_l1 = normalize_l1(X)\n",
    "# X_norm_l2 = normalize_l2(X)\n",
    "# fig, (ax1, ax2, ax3) = plt.subplots(3, 1, figsize=(5, 10))\n",
    "\n",
    "# ax1.hist(X, bins=60, alpha=0.5, histtype='step', label='X')\n",
    "# ax2.hist(X_norm_l1, bins=60, alpha=0.5, histtype='step', label='X_norm_l1')\n",
    "# ax3.hist(X_norm_l2, bins=60, alpha=0.5, histtype='step', label='X_norm_l2')\n",
    "# plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>avg_age</th>\n",
       "      <th>avg_age_l1</th>\n",
       "      <th>avg_age_l2</th>\n",
       "      <th>avg_age_01</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>0.000831</td>\n",
       "      <td>0.022735</td>\n",
       "      <td>0.271174</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>38.0</td>\n",
       "      <td>0.001436</td>\n",
       "      <td>0.039270</td>\n",
       "      <td>0.472229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0.000983</td>\n",
       "      <td>0.026869</td>\n",
       "      <td>0.321438</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0.001323</td>\n",
       "      <td>0.036170</td>\n",
       "      <td>0.434531</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0.001323</td>\n",
       "      <td>0.036170</td>\n",
       "      <td>0.434531</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   avg_age  avg_age_l1  avg_age_l2  avg_age_01\n",
       "0     22.0    0.000831    0.022735    0.271174\n",
       "1     38.0    0.001436    0.039270    0.472229\n",
       "2     26.0    0.000983    0.026869    0.321438\n",
       "3     35.0    0.001323    0.036170    0.434531\n",
       "4     35.0    0.001323    0.036170    0.434531"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "avg_age_l1 = normalize_l1(df['avg_age'].values)\n",
    "avg_age_l2 = normalize_l2(df['avg_age'].values)\n",
    "avg_age_01 = normalize_to_01(df['avg_age'].values)\n",
    "\n",
    "df['avg_age_l1'] = avg_age_l1\n",
    "df['avg_age_l2'] = avg_age_l2\n",
    "df['avg_age_01'] = avg_age_01\n",
    "\n",
    "df[['avg_age', 'avg_age_l1', 'avg_age_l2', 'avg_age_01']].head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2, figsize=(10, 10))\n",
    "df['avg_age'].plot(kind='hist', ax=ax1)\n",
    "ax1.set_title('avg_age')\n",
    "df['avg_age_l1'].plot(kind='hist', ax=ax2)\n",
    "ax2.set_title('avg_age_l1')\n",
    "df['avg_age_l2'].plot(kind='hist', ax=ax3)\n",
    "ax3.set_title('avg_age_l2')\n",
    "df['avg_age_01'].plot(kind='hist', ax=ax4)\n",
    "ax4.set_title('avg_age_01')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# X = df[['avg_age', 'avg_age_l1']].values\n",
    "plt.scatter(df['avg_age'], df['avg_age_l1'], label='avg_age_l1')\n",
    "plt.scatter(df['avg_age'], df['avg_age_l2'], label='avg_age_l2')\n",
    "plt.scatter(df['avg_age'], df['avg_age_01'], label='avg_age_01')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cabin - NLP category"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 整理出每一個Cabin的個數並排序(hint:Counter(), sorted())\n",
    "# 1. 使用Counter計算個別item出現的次數\n",
    "# 2. 使用items上dict變成list(可排序)\n",
    "# 3. sorted 對list中每一item的第一個ELEMENT進行排序\n",
    "# 4. 設定reverse參數\n",
    "#=============your works starts===============#\n",
    "sorted_cabin_counter = sorted(Counter(df['Cabin'].astype(str)).items(), key=lambda x:x[1], reverse=True)\n",
    "#==============your works ends================#\n",
    "\n",
    "\n",
    "print(sorted_cabin_counter[:10])\n",
    "# [('nan', 687), ('G6', 4), ('C23 C25 C27', 4), ('B96 B98', 4), ('F33', 3), ('E101', 3), ('F2', 3), ('D', 3), ('C22 C26', 3), ('C123', 2)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{0: 'A', 1: 'B', 2: 'C'}\n"
     ]
    }
   ],
   "source": [
    "test_arr = [\"A\", \"B\", \"C\"]\n",
    "\n",
    "# enumerate的用法: 替test_arr寫上編號，並轉換成dict型別\n",
    "#=============your works starts===============#\n",
    "enumerate_res = dict([(idx, value) for idx, value in enumerate(test_arr)])\n",
    "#==============your works ends================#\n",
    "\n",
    "print(enumerate_res)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# test_arr = pd.Series([\"A\", \"B\", \"C\", \"A\", \"D\", \"F\", \"B\"])\n",
    "# test_dict = dict([(value, key) for key, value in list(enumerate(sorted(set(test_arr))))])\n",
    "# test_arr.apply(test_dict.get)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['n' 'c' 'n' 'c' 'n' 'n' 'e' 'n' 'n' 'n']\n",
      "{'c': 0, 'n': 1, 'g': 2, 'b': 3, 't': 4, 'e': 5, 'd': 6, 'f': 7, 'a': 8}\n",
      "[1 0 1 0 1 1 5 1 1 1]\n"
     ]
    }
   ],
   "source": [
    "# 抓出第一個char出來分類，並轉成int類別(hint: enumerate)\n",
    "# new_Cabin: 抓出第一個char，轉str，轉小寫 (1.轉換成str型別 2.用apply找到cabin中的第一個char並轉小寫 3.指派到新的變數)\n",
    "# mapping_dict: 找到每個開頭char所屬的編號 (1.取不重複的 2.轉成可排序的 3.排序 4.列舉 5.用for loop 展開generator 6.把key跟value對調 7.轉換成dict型別 8.指派到新的變數)\n",
    "# new_Cabin_int: 將char轉換成編號 (1.測試mapping_dict的索引功能('a' => 0) 2.用new_Cabin的apply去索引其編號 (lambda, mapping_dict) 3.用values轉換成np.array()型別)\n",
    "#=============your works starts===============#\n",
    "# new_Cabin = [cabin[0].lower() for cabin in df['Cabin'].astype(str)]\n",
    "new_Cabin = df['Cabin'].astype(str).apply(lambda x:x[0].lower())\n",
    "mapping_dict = dict([(value, key) for (key, value) in enumerate(sorted(list(set(new_Cabin))))])\n",
    "new_Cabin_int = new_Cabin.apply(lambda x:mapping_dict[x]).values\n",
    "#==============your works ends================#\n",
    "\n",
    "\n",
    "print(new_Cabin.values[:10])\n",
    "print(mapping_dict)\n",
    "print(new_Cabin_int[:10])\n",
    "# ['n' 'c' 'n' 'c' 'n' 'n' 'e' 'n' 'n' 'n']\n",
    "# {'e': 0, 'f': 1, 'd': 2, 't': 3, 'c': 4, 'n': 5, 'b': 6, 'g': 7, 'a': 8}\n",
    "# [5 4 5 4 5 5 0 5 5 5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Cabin</th>\n",
       "      <th>cabin_cat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>C85</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>C123</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Cabin  cabin_cat\n",
       "0   NaN          1\n",
       "1   C85          0\n",
       "2   NaN          1\n",
       "3  C123          0\n",
       "4   NaN          1"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['cabin_cat'] = new_Cabin_int\n",
    "df[['Cabin', 'cabin_cat']].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Sex - Category"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Counter({'male': 577, 'female': 314})\n"
     ]
    }
   ],
   "source": [
    "# 請算出Sex共有幾個類別，每一個類別共出現幾次(hint:Counter)\n",
    "#=============your works starts===============#\n",
    "counter = Counter(df['Sex'])\n",
    "#==============your works ends================#\n",
    "\n",
    "\n",
    "print(counter)\n",
    "#Counter({'male': 577, 'female': 314})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Counter(sex_cat) Counter({0: 577, 1: 314})\n"
     ]
    }
   ],
   "source": [
    "# 創造出一個與df['Sex']等長的array，並將df['Sex']中的male換成1，female換成0\n",
    "#=============your works starts===============#\n",
    "sex_mapping = dict([(v, k) for k, v in enumerate(set(df['Sex'].astype(str)))])\n",
    "sex_cat = df['Sex'].apply(sex_mapping.get)\n",
    "#==============your works ends================#\n",
    "\n",
    "\n",
    "print(\"Counter(sex_cat)\", Counter(sex_cat))\n",
    "#Counter(sex_cat) Counter({1: 577, 0: 314})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Counter({0: 577, 1: 314})"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['sex_cat'] = sex_cat\n",
    "Counter(df['sex_cat'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Ticket - Category"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[('A/5 21171', 1), ('PC 17599', 1), ('STON/O2. 3101282', 1), ('373450', 1), ('330877', 1), ('17463', 1), ('113783', 1), ('A/5. 2151', 1), ('350406', 1), ('248706', 1), ('244373', 1), ('345763', 1), ('2649', 1), ('248698', 1), ('330923', 1), ('113788', 1), ('2631', 1), ('330959', 1), ('349216', 1), ('PC 17601', 1), ('335677', 1), ('C.A. 24579', 1), ('2677', 1), ('A./5. 2152', 1), ('7546', 1), ('349253', 1), ('330958', 1), ('S.C./A.4. 23567', 1), ('370371', 1), ('14311', 1), ('2662', 1), ('A/4. 39886', 1), ('2926', 1), ('113509', 1), ('19947', 1), ('C.A. 31026', 1), ('2697', 1), ('2669', 1), ('PC 17605', 1), ('C.A. 29395', 1), ('S.P. 3464', 1), ('3101281', 1), ('315151', 1), ('C.A. 33111', 1), ('2680', 1), ('348123', 1), ('349208', 1), ('374746', 1), ('345767', 1), ('345779', 1), ('330932', 1), ('113059', 1), ('SO/C 14885', 1), ('SOTON/OQ 392086', 1), ('343275', 1), ('343276', 1), ('347466', 1), ('W.E.P. 5734', 1), ('364500', 1), ('374910', 1), ('PC 17754', 1), ('PC 17759', 1), ('349245', 1), ('349215', 1), ('7540', 1), ('3101276', 1), ('349207', 1), ('343120', 1), ('312991', 1), ('349249', 1), ('324669', 1), ('4136', 1), ('STON/O 2. 3101294', 1), ('370369', 1), ('A4. 54510', 1), ('27267', 1), ('370372', 1), ('C 17369', 1), ('347061', 1), ('349241', 1), ('SOTON/O.Q. 3101307', 1), ('A/5. 3337', 1), ('228414', 1), ('C.A. 29178', 1), ('SC/PARIS 2133', 1), ('11752', 1), ('347081', 1), ('365222', 1), ('231945', 1), ('350043', 1), ('244310', 1), ('S.O.P. 1166', 1), ('A.5. 11206', 1), ('A/5. 851', 1), ('Fa 265302', 1), ('PC 17597', 1), ('35851', 1), ('SOTON/OQ 392090', 1), ('315037', 1), ('371362', 1), ('C.A. 33595', 1), ('347068', 1), ('315093', 1), ('PC 17318', 1), ('111240', 1), ('STON/O 2. 3101280', 1), ('17764', 1), ('350404', 1), ('PC 17595', 1), ('250653', 1), ('SC/PARIS 2131', 1), ('315153', 1), ('113767', 1), ('111428', 1), ('349247', 1), ('234604', 1), ('28424', 1), ('350046', 1), ('PC 17610', 1), ('368703', 1), ('4579', 1), ('370370', 1), ('248747', 1), ('345770', 1), ('3101264', 1), ('2628', 1), ('A/5 3540', 1), ('367231', 1), ('112277', 1), ('SOTON/O.Q. 3101311', 1), ('F.C.C. 13528', 1), ('A/5 21174', 1), ('250646', 1), ('367229', 1), ('STON/O2. 3101283', 1), ('11813', 1), ('W/C 14208', 1), ('SOTON/OQ 392089', 1), ('220367', 1), ('21440', 1), ('349234', 1), ('PP 4348', 1), ('SW/PP 751', 1), ('A/5 21173', 1), ('236171', 1), ('347067', 1), ('237442', 1), ('C.A. 29566', 1), ('W./C. 6609', 1), ('26707', 1), ('28665', 1), ('SCO/W 1585', 1), ('W./C. 14263', 1), ('STON/O 2. 3101275', 1), ('2694', 1), ('347071', 1), ('362316', 1), ('113514', 1), ('2650', 1), ('PC 17585', 1), ('384461', 1), ('112059', 1), ('382649', 1), ('C.A. 17248', 1), ('347083', 1), ('PC 17596', 1), ('370375', 1), ('347073', 1), ('336439', 1), ('347464', 1), ('345778', 1), ('A/5. 10482', 1), ('113056', 1), ('349239', 1), ('345774', 1), ('349206', 1), ('237798', 1), ('370373', 1), ('SC/Paris 2163', 1), ('349236', 1), ('349233', 1), ('PC 17612', 1), ('2693', 1), ('19988', 1), ('9234', 1), ('226593', 1), ('A/5 2466', 1), ('250651', 1), ('349243', 1), ('347470', 1), ('29011', 1), ('A/5 21172', 1), ('349219', 1), ('234818', 1), ('345364', 1), ('28551', 1), ('113043', 1), ('349225', 1), ('7598', 1), ('113784', 1), ('248740', 1), ('244361', 1), ('229236', 1), ('248733', 1), ('31418', 1), ('386525', 1), ('315088', 1), ('7267', 1), ('113510', 1), ('2695', 1), ('2647', 1), ('345783', 1), ('237671', 1), ('330931', 1), ('330980', 1), ('SC/PARIS 2167', 1), ('SOTON/O.Q. 3101310', 1), ('C 7076', 1), ('110813', 1), ('2626', 1), ('14313', 1), ('11765', 1), ('3101267', 1), ('323951', 1), ('C 7077', 1), ('113503', 1), ('2648', 1), ('347069', 1), ('STON/O 2. 3101293', 1), ('349227', 1), ('27849', 1), ('367655', 1), ('SC 1748', 1), ('350034', 1), ('3101277', 1), ('350052', 1), ('350407', 1), ('244278', 1), ('240929', 1), ('STON/O 2. 3101289', 1), ('341826', 1), ('4137', 1), ('315096', 1), ('28664', 1), ('347064', 1), ('312992', 1), ('349222', 1), ('394140', 1), ('STON/O 2. 3101269', 1), ('343095', 1), ('28220', 1), ('250652', 1), ('28228', 1), ('349254', 1), ('A/5. 13032', 1), ('315082', 1), ('A/4. 34244', 1), ('2003', 1), ('364851', 1), ('SOTON/O.Q. 392078', 1), ('110564', 1), ('SC/AH 3085', 1), ('STON/O 2. 3101274', 1), ('C.A. 18723', 1), ('345769', 1), ('347076', 1), ('230434', 1), ('65306', 1), ('33638', 1), ('113794', 1), ('113786', 1), ('65303', 1), ('113051', 1), ('A/5 2817', 1), ('349240', 1), ('13509', 1), ('17464', 1), ('F.C.C. 13531', 1), ('371060', 1), ('19952', 1), ('364506', 1), ('111320', 1), ('234360', 1), ('A/S 2816', 1), ('SOTON/O.Q. 3101306', 1), ('113792', 1), ('36209', 1), ('323592', 1), ('315089', 1), ('SC/AH Basle 541', 1), ('7553', 1), ('3460', 1), ('350060', 1), ('3101298', 1), ('239854', 1), ('A/5 3594', 1), ('4134', 1), ('11771', 1), ('A.5. 18509', 1), ('65304', 1), ('SOTON/OQ 3101317', 1), ('113787', 1), ('PC 17609', 1), ('A/4 45380', 1), ('C.A. 6212', 1), ('350035', 1), ('315086', 1), ('364846', 1), ('330909', 1), ('4135', 1), ('111427', 1), ('C 4001', 1), ('382651', 1), ('SOTON/OQ 3101316', 1), ('PC 17473', 1), ('PC 17603', 1), ('349209', 1), ('36967', 1), ('C.A. 34260', 1), ('226875', 1), ('349242', 1), ('349252', 1), ('2624', 1), ('2700', 1), ('367232', 1), ('W./C. 14258', 1), ('PC 17483', 1), ('3101296', 1), ('29104', 1), ('2641', 1), ('2690', 1), ('315084', 1), ('113050', 1), ('364498', 1), ('13568', 1), ('693', 1), ('SC/PARIS 2146', 1), ('244358', 1), ('330979', 1), ('2620', 1), ('347085', 1), ('113807', 1), ('11755', 1), ('345572', 1), ('372622', 1), ('349251', 1), ('218629', 1), ('SOTON/OQ 392082', 1), ('SOTON/O.Q. 392087', 1), ('349205', 1), ('2686', 1), ('350417', 1), ('S.W./PP 752', 1), ('11769', 1), ('PC 17474', 1), ('14312', 1), ('A/4. 20589', 1), ('243880', 1), ('2689', 1), ('STON/O 2. 3101286', 1), ('237789', 1), ('13049', 1), ('3411', 1), ('237565', 1), ('13567', 1), ('14973', 1), ('A./5. 3235', 1), ('STON/O 2. 3101273', 1), ('A/5 3902', 1), ('364848', 1), ('SC/AH 29037', 1), ('2664', 1), ('349214', 1), ('113796', 1), ('364511', 1), ('111426', 1), ('349910', 1), ('349246', 1), ('113804', 1), ('SOTON/O.Q. 3101305', 1), ('370377', 1), ('364512', 1), ('31028', 1), ('11753', 1), ('350029', 1), ('36963', 1), ('219533', 1), ('349224', 1), ('334912', 1), ('27042', 1), ('347743', 1), ('13214', 1), ('112052', 1), ('237668', 1), ('STON/O 2. 3101292', 1), ('350050', 1), ('349231', 1), ('13213', 1), ('S.O./P.P. 751', 1), ('CA. 2314', 1), ('349221', 1), ('8475', 1), ('330919', 1), ('365226', 1), ('349223', 1), ('29751', 1), ('2623', 1), ('5727', 1), ('349210', 1), ('STON/O 2. 3101285', 1), ('234686', 1), ('312993', 1), ('A/5 3536', 1), ('F.C. 12750', 1), ('C.A. 24580', 1), ('244270', 1), ('239856', 1), ('349912', 1), ('342826', 1), ('4138', 1), ('330935', 1), ('6563', 1), ('349228', 1), ('350036', 1), ('349256', 1), ('2672', 1), ('113800', 1), ('248731', 1), ('363592', 1), ('35852', 1), ('348121', 1), ('PC 17475', 1), ('36864', 1), ('350025', 1), ('223596', 1), ('PC 17476', 1), ('PC 17482', 1), ('113028', 1), ('7545', 1), ('348124', 1), ('34218', 1), ('36568', 1), ('347062', 1), ('350048', 1), ('12233', 1), ('250643', 1), ('315094', 1), ('36866', 1), ('236853', 1), ('STON/O2. 3101271', 1), ('239855', 1), ('28425', 1), ('233639', 1), ('349201', 1), ('349218', 1), ('16988', 1), ('376566', 1), ('STON/O 2. 3101288', 1), ('250648', 1), ('113773', 1), ('335097', 1), ('29103', 1), ('345780', 1), ('349204', 1), ('350042', 1), ('29108', 1), ('363294', 1), ('SOTON/O2 3101272', 1), ('2663', 1), ('347074', 1), ('112379', 1), ('364850', 1), ('8471', 1), ('345781', 1), ('350047', 1), ('2674', 1), ('29105', 1), ('347078', 1), ('383121', 1), ('36865', 1), ('2687', 1), ('113501', 1), ('SOTON/O.Q. 3101312', 1), ('374887', 1), ('3101265', 1), ('12460', 1), ('PC 17600', 1), ('349203', 1), ('28213', 1), ('17465', 1), ('349244', 1), ('2685', 1), ('2625', 1), ('347089', 1), ('347063', 1), ('112050', 1), ('347087', 1), ('248723', 1), ('3474', 1), ('28206', 1), ('364499', 1), ('112058', 1), ('STON/O2. 3101290', 1), ('C 7075', 1), ('315098', 1), ('19972', 1), ('368323', 1), ('367228', 1), ('2671', 1), ('347468', 1), ('2223', 1), ('PC 17756', 1), ('315097', 1), ('392092', 1), ('11774', 1), ('SOTON/O2 3101287', 1), ('2683', 1), ('315090', 1), ('C.A. 5547', 1), ('349213', 1), ('347060', 1), ('PC 17592', 1), ('392091', 1), ('113055', 1), ('2629', 1), ('350026', 1), ('28134', 1), ('17466', 1), ('233866', 1), ('236852', 1), ('SC/PARIS 2149', 1), ('PC 17590', 1), ('345777', 1), ('349248', 1), ('695', 1), ('345765', 1), ('2667', 1), ('349212', 1), ('349217', 1), ('349257', 1), ('7552', 1), ('C.A./SOTON 34068', 1), ('SOTON/OQ 392076', 1), ('211536', 1), ('112053', 1), ('111369', 1), ('370376', 1), ('113803', 2), ('237736', 2), ('PP 9549', 2), ('239865', 2), ('PC 17569', 2), ('PC 17604', 2), ('113789', 2), ('345764', 2), ('2651', 2), ('11668', 2), ('349237', 2), ('113572', 2), ('36973', 2), ('2661', 2), ('248738', 2), ('364516', 2), ('3101278', 2), ('C.A. 2315', 2), ('231919', 2), ('244367', 2), ('35281', 2), ('110465', 2), ('2665', 2), ('2627', 2), ('PC 17558', 2), ('2668', 2), ('7534', 2), ('PC 17593', 2), ('2678', 2), ('STON/O2. 3101279', 2), ('C.A. 33112', 2), ('113776', 2), ('113505', 2), ('230136', 2), ('370365', 2), ('364849', 2), ('347054', 2), ('2699', 2), ('243847', 2), ('19943', 2), ('367230', 2), ('19928', 2), ('250649', 2), ('11751', 2), ('244252', 2), ('A/5. 3336', 2), ('370129', 2), ('230433', 2), ('113798', 2), ('250644', 2), ('C.A. 2673', 2), ('19877', 2), ('11967', 2), ('367226', 2), ('PC 17758', 2), ('P/PP 3381', 2), ('PC 17485', 2), ('11767', 2), ('PC 17608', 2), ('36928', 2), ('16966', 2), ('111361', 2), ('PC 17611', 2), ('C.A. 37671', 2), ('2691', 2), ('PC 17477', 2), ('2653', 2), ('28403', 2), ('347080', 2), ('250655', 2), ('376564', 2), ('13507', 2), ('17453', 2), ('31027', 2), ('36947', 2), ('26360', 2), ('12749', 2), ('PC 17761', 2), ('WE/P 5735', 2), ('2908', 2), ('A/4 48871', 2), ('358585', 2), ('220845', 2), ('2659', 2), ('54636', 2), ('19996', 2), ('29750', 2), ('17474', 2), ('250647', 2), ('113806', 2), ('392096', 2), ('S.O./P.P. 3', 2), ('W./C. 6607', 2), ('S.C./PARIS 2079', 2), ('347742', 3), ('SC/Paris 2123', 3), ('PC 17572', 3), ('C.A. 34651', 3), ('371110', 3), ('230080', 3), ('363291', 3), ('35273', 3), ('C.A. 31921', 3), ('110152', 3), ('PC 17755', 3), ('110413', 3), ('PC 17582', 3), ('PC 17760', 3), ('13502', 3), ('239853', 3), ('F.C.C. 13529', 3), ('29106', 3), ('345773', 3), ('248727', 3), ('24160', 3), ('349909', 4), ('347077', 4), ('19950', 4), ('W./C. 6608', 4), ('4133', 4), ('LINE', 4), ('113781', 4), ('17421', 4), ('PC 17757', 4), ('113760', 4), ('2666', 4), ('382652', 5), ('S.O.C. 14879', 5), ('3101295', 6), ('CA 2144', 6), ('347088', 6), ('347082', 7), ('1601', 7), ('CA. 2343', 7)]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# 整理出每一個Ticket的個數並排序\n",
    "# 1. 用value_counts計算每一item出現的個數\n",
    "# 2. 用items轉成tuple list(generator)型別\n",
    "# 3. 用list把generator展開\n",
    "# 4. 指派到新的變數\n",
    "#=============your works starts===============#\n",
    "sorted_ticket_counter = list(df['Ticket'].value_counts().items())\n",
    "#==============your works ends================#\n",
    "\n",
    "print(sorted_ticket_counter)\n",
    "# [('110152', 3), ('110413', 3), ('110465', 2), ('110564', 1), ('110813', 1), ('111240', 1), ('111320', 1), ('111361', 2), ('111369', 1),"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Counter({1: 661, 5: 65, 3: 47, 8: 30, 2: 29, 7: 18, 6: 17, 9: 13, 4: 7, 0: 4})\n"
     ]
    }
   ],
   "source": [
    "# ticket\n",
    "ticket_cat = {}\n",
    "for ticket in df['Ticket']:\n",
    "    if ticket.isdigit():\n",
    "        ticket_cat[ticket] = 1\n",
    "    elif ticket.startswith('A'):\n",
    "        ticket_cat[ticket] = 2\n",
    "    elif ticket.startswith('C'):\n",
    "        ticket_cat[ticket] = 3\n",
    "    elif ticket.startswith('F'):\n",
    "        ticket_cat[ticket] = 4\n",
    "    elif ticket.startswith('P'):\n",
    "        ticket_cat[ticket] = 5\n",
    "    elif ticket.startswith('SOTON'):\n",
    "        ticket_cat[ticket] = 6\n",
    "    elif ticket.startswith('STON'):\n",
    "        ticket_cat[ticket] = 7\n",
    "    elif ticket.startswith('S'):\n",
    "        ticket_cat[ticket] = 8\n",
    "    elif ticket.startswith('W'):\n",
    "        ticket_cat[ticket] = 9\n",
    "    else:\n",
    "        ticket_cat[ticket] = 0\n",
    "df['ticket_cat'] = df['Ticket'].apply(ticket_cat.get)\n",
    "print(Counter(df['ticket_cat']))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Embarked - Category"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[('C', 168), ('Q', 77), ('S', 644), ('nan', 2)]\n"
     ]
    }
   ],
   "source": [
    "# 整理出每一個Embarked的個數並排序(hint:Counter(), sorted())\n",
    "#=============your works starts===============#\n",
    "sorted_embarked_counter = sorted(dict(Counter(df['Embarked'].astype(str))).items(), key=lambda x:x[0])\n",
    "#==============your works ends================#\n",
    "\n",
    "print(sorted_embarked_counter)\n",
    "# [('C', 168), ('Q', 77), ('S', 644), ('nan', 2)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{nan: 0, 'S': 1, 'C': 2, 'Q': 3}\n"
     ]
    }
   ],
   "source": [
    "# 創造embarked的類別對應dict\n",
    "#=============your works starts===============#\n",
    "embarked_mapping = dict([(e,i)for i, e in enumerate(set(df['Embarked']))])\n",
    "#==============your works ends================#\n",
    "\n",
    "\n",
    "print(embarked_mapping)\n",
    "#{nan: 0, 'S': 1, 'Q': 2, 'C': 3}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Counter({1: 644, 2: 168, 3: 77, 0: 2})\n"
     ]
    }
   ],
   "source": [
    "# 轉換embarked為數字類別\n",
    "#=============your works starts===============#\n",
    "df['embarked_cat'] = df['Embarked'].apply(embarked_mapping.get)\n",
    "#==============your works ends================#\n",
    "\n",
    "print(Counter(df['embarked_cat']))\n",
    "#Counter({1: 644, 3: 168, 2: 77, 0: 2})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "dict[key] v.s. dict.get(key)\n",
    "    - get 可以設定第二個參數，找不到這個key要回傳的值，default=None\n",
    "    - []索引如果找不到key，會跳error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# X = np.array([1, np.nan, None]).astype(float)\n",
    "# X == None\n",
    "# np.isnan(X)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Name - NLP category"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['(Florence Briggs Thayer)']\n",
      "['(Dother Thier)']\n"
     ]
    }
   ],
   "source": [
    "# re.findall()的使用方法\n",
    "test_str = 'Cumings, Mrs. John Bradley (Florence Briggs Thayer)'\n",
    "test_str1 = 'St. Clumbia (Dother Thier)'\n",
    "\n",
    "# 請取出「(」、「)」中的文字\n",
    "#=============your works starts===============#\n",
    "pattern = \"\\(.+?\\)\"\n",
    "sub_str = re.findall(pattern, test_str)\n",
    "sub_str1 = re.findall(pattern, test_str1)\n",
    "#==============your works ends================#\n",
    "\n",
    "print(sub_str)\n",
    "print(sub_str1)\n",
    "# ['(Florence Briggs Thayer)']\n",
    "# ['(Dother Thier)']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Mr' 'Mrs' 'Miss' 'Mrs' 'Mr' 'Mr' 'Mr' 'Master' 'Mrs' 'Mrs']\n"
     ]
    }
   ],
   "source": [
    "# 請找到位在\",\"以及\".\"之間的所有字並將\" \", \".\", \",\"去掉(hint: re.findall(), str.replace())\n",
    "# 1. 用re.findall找到所有在\",\"以及\".\"的詞\n",
    "# 2. 所引出第零的詞\n",
    "# 3. 用replace拿掉\",\", \".\"以及\" \"\n",
    "def find_call(name):\n",
    "    #=============your works starts===============#\n",
    "    name = re.findall('\\,.+\\.', name)[0].replace(\",\", \"\").replace(\".\", \"\").replace(\" \", \"\")\n",
    "    #==============your works ends================#\n",
    "    return name \n",
    "\n",
    "title_cat_series = df['Name'].apply(find_call)\n",
    "print(title_cat_series.values[:10])\n",
    "#['Mr' 'Mrs' 'Miss' 'Mrs' 'Mr' 'Mr' 'Mr' 'Master' 'Mrs' 'Mrs']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Counter({1: 517, 2: 185, 3: 126, 4: 40, 5: 23})\n"
     ]
    }
   ],
   "source": [
    "title_mapping= {\n",
    "    'Ms':\"Miss\",\n",
    "    'Mlle':\"Miss\",\n",
    "    'Miss':\"Miss\",\n",
    "    'Mrs':\"Mrs\",\n",
    "    'Mme':\"Mrs\",\n",
    "    'MrsMartin(ElizabethL':\"Mrs\",\n",
    "    'Mr':\"Mr\"\n",
    "    }\n",
    "\n",
    "title_cat = {\"Mr\": 1, \"Miss\": 2, \"Mrs\": 3, \"Master\": 4, \"Rare\": 5}\n",
    "\n",
    "def process_title(call):\n",
    "    if title_cat.get(call):\n",
    "        return title_cat.get(call)\n",
    "    else:\n",
    "        return title_cat.get(title_mapping.get(call, \"Rare\"))\n",
    "\n",
    "df['title_cat'] = title_cat_series.apply(process_title) \n",
    "print(Counter(df['title_cat']))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Title - Length"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Counter({1: 303, 2: 237, 0: 204, 3: 57, 4: 53, 5: 26, 6: 8, 7: 2, 9: 1})\n"
     ]
    }
   ],
   "source": [
    "# 算出df['Name']中每一個名字的長度並放進一個array\n",
    "#=============your works starts===============#\n",
    "name_length = df['Name'].apply(len)\n",
    "#==============your works ends================#\n",
    "\n",
    "\n",
    "print(Counter(pd.cut(name_length, bins=10, labels=range(10))))\n",
    "# Counter({1: 303, 2: 237, 0: 204, 3: 57, 4: 53, 5: 26, 6: 8, 7: 2, 9: 1})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['name_length'] = name_length"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\users\\thinktron\\appdata\\local\\programs\\python\\python36\\lib\\site-packages\\ipykernel_launcher.py:3: FutureWarning: `Series.plot()` should not be called with positional arguments, only keyword arguments. The order of positional arguments will change in the future. Use `Series.plot(kind='hist')` instead of `Series.plot('hist',)`.\n",
      "  This is separate from the ipykernel package so we can avoid doing imports until\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 劃出每一長度區間次數的長條分布圖(如長度界在10~20之間的有出現約150次)(hint: df[col].plot('hist'))\n",
    "#!=============your works starts===============!#\n",
    "df['name_length'].plot('hist')\n",
    "#!==============your works ends================!#\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fare - PCA, smooth noisy data, feature generation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['PassengerId', 'Survived', 'Pclass', 'Age', 'SibSp', 'Parch', 'Fare', 'avg_age', 'median_age', 'random_age', 'avg_age_l1', 'avg_age_l2', 'avg_age_01', 'name_length']\n"
     ]
    }
   ],
   "source": [
    "# 請找出dtype是np.int64或np.float64且名稱不以'_cat'結尾的欄位。\n",
    "#=============your works starts===============#\n",
    "number_cols = [col for col in df.columns if df[col].dtype in [np.int64, np.float64] and not col.endswith('_cat')]\n",
    "#==============your works ends================#\n",
    "\n",
    "\n",
    "print(number_cols)\n",
    "# ['PassengerId', 'Survived', 'Pclass', 'Age', 'SibSp', 'Parch', 'Fare', 'avg_age', 'median_age', 'random_age', 'avg_age_l1', 'avg_age_l2', 'name_length']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "請參照[這個連結](http://scikit-learn.org/stable/auto_examples/datasets/plot_iris_dataset.html#sphx-glr-auto-examples-datasets-plot-iris-dataset-py)，劃出以下這張PCA圖\n",
    "\n",
    "![PCA_chart](./imgs/PCA.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X = df[['Pclass', 'Parch', 'SibSp', 'Fare', 'avg_age', 'name_length']].values\n",
    "Y = np.array(df['Survived'])\n",
    "\n",
    "#!=============your works starts===============!#\n",
    "fig = plt.figure(1, figsize=(8, 6))\n",
    "ax = Axes3D(fig, elev=-150, azim=110)\n",
    "\n",
    "X_reduced = PCA(n_components=3).fit_transform(X)\n",
    "ax.scatter(X_reduced[:, 0], X_reduced[:, 1], X_reduced[:, 2], c=Y,\n",
    "           cmap=plt.cm.Set1, edgecolor='k', s=40)\n",
    "ax.set_title(\"PCA with non-log(Fare)\")\n",
    "ax.set_xlabel(\"1st eigenvector\")\n",
    "ax.w_xaxis.set_ticklabels([])\n",
    "ax.set_ylabel(\"2nd eigenvector\")\n",
    "ax.w_yaxis.set_ticklabels([])\n",
    "ax.set_zlabel(\"3rd eigenvector\")\n",
    "ax.w_zaxis.set_ticklabels([])\n",
    "#!==============your works ends================!#\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Parch</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Fare</th>\n",
       "      <th>avg_age</th>\n",
       "      <th>name_length</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>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>2.308642</td>\n",
       "      <td>0.381594</td>\n",
       "      <td>0.523008</td>\n",
       "      <td>32.204208</td>\n",
       "      <td>29.699118</td>\n",
       "      <td>26.965208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.836071</td>\n",
       "      <td>0.806057</td>\n",
       "      <td>1.102743</td>\n",
       "      <td>49.693429</td>\n",
       "      <td>13.002015</td>\n",
       "      <td>9.281607</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>0.000000</td>\n",
       "      <td>0.420000</td>\n",
       "      <td>12.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7.910400</td>\n",
       "      <td>22.000000</td>\n",
       "      <td>20.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>14.454200</td>\n",
       "      <td>29.699118</td>\n",
       "      <td>25.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>30.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>512.329200</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>82.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Pclass       Parch       SibSp        Fare     avg_age  name_length\n",
       "count  891.000000  891.000000  891.000000  891.000000  891.000000   891.000000\n",
       "mean     2.308642    0.381594    0.523008   32.204208   29.699118    26.965208\n",
       "std      0.836071    0.806057    1.102743   49.693429   13.002015     9.281607\n",
       "min      1.000000    0.000000    0.000000    0.000000    0.420000    12.000000\n",
       "25%      2.000000    0.000000    0.000000    7.910400   22.000000    20.000000\n",
       "50%      3.000000    0.000000    0.000000   14.454200   29.699118    25.000000\n",
       "75%      3.000000    0.000000    1.000000   31.000000   35.000000    30.000000\n",
       "max      3.000000    6.000000    8.000000  512.329200   80.000000    82.000000"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 請找出標準差最大的欄位['Pclass', 'Parch', 'SibSp', 'Fare', 'avg_age', 'name_length']\n",
    "#!=============your works starts===============!#\n",
    "df[['Pclass', 'Parch', 'SibSp', 'Fare', 'avg_age', 'name_length']].describe()\n",
    "#!==============your works ends================!#\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # 請找出Fare的平均值(mean)，並填入df['Fare']中\n",
    "# #=============your works starts===============#\n",
    "# avg_fare = df['Fare'].mean()\n",
    "# df[pd.isnull(df['Fare'])] = avg_fare\n",
    "# #==============your works ends================#\n",
    "# # df['Fare'].fillna(avg_fare)\n",
    "\n",
    "\n",
    "# print(\"avg_fare\", avg_fare)\n",
    "# # 32.204207968574636\n",
    "# print(\"number of null of Fare:\", len(df[pd.isnull(df['Fare'])]))\n",
    "# # number of null of Fare: 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number of Fare equals zero: 0\n"
     ]
    }
   ],
   "source": [
    "# 找出Fare==0的row，補上Fare=1\n",
    "#!=============your works starts===============!#\n",
    "df.loc[df['Fare']==0, 'Fare'] = 1\n",
    "#!==============your works ends================!#\n",
    "\n",
    "print(\"number of Fare equals zero:\", len(df[df['Fare']==0]))\n",
    "# number of Fare equals zero: 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.86033801 1.8529878  0.89899927 1.72509452 0.90579588]\n"
     ]
    }
   ],
   "source": [
    "# 請算出Fare以10為底的log值\n",
    "#=============your works starts===============#\n",
    "fare_log10 = np.log10(df['Fare'].values)\n",
    "#==============your works ends================#\n",
    "\n",
    "\n",
    "print(fare_log10[:5])\n",
    "# [0.86033801 1.8529878  0.89899927 1.72509452 0.90579588]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Fare</th>\n",
       "      <th>fare_log10</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>7.2500</td>\n",
       "      <td>0.860338</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>71.2833</td>\n",
       "      <td>1.852988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>7.9250</td>\n",
       "      <td>0.898999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>53.1000</td>\n",
       "      <td>1.725095</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>8.0500</td>\n",
       "      <td>0.905796</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Fare  fare_log10\n",
       "0   7.2500    0.860338\n",
       "1  71.2833    1.852988\n",
       "2   7.9250    0.898999\n",
       "3  53.1000    1.725095\n",
       "4   8.0500    0.905796"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['fare_log10'] = fare_log10\n",
    "df[['Fare', 'fare_log10']].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    891.000000\n",
       "mean       1.256781\n",
       "std        0.435553\n",
       "min        0.000000\n",
       "25%        0.898198\n",
       "50%        1.159994\n",
       "75%        1.491362\n",
       "max        2.709549\n",
       "Name: fare_log10, dtype: float64"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['fare_log10'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X = np.matrix(df[['Parch', 'SibSp', 'avg_age', 'fare_log10']])\n",
    "Y = np.array(df['Survived'])\n",
    "\n",
    "fig = plt.figure(1, figsize=(8, 6))\n",
    "ax = Axes3D(fig, elev=-150, azim=110)\n",
    "\n",
    "X_reduced = PCA(n_components=3).fit_transform(X)\n",
    "ax.scatter(X_reduced[:, 0], X_reduced[:, 1], X_reduced[:, 2], c=Y,\n",
    "           cmap=plt.cm.Set1, edgecolor='k', s=40)\n",
    "ax.set_title(\"PCA with non-log(Fare)\")\n",
    "ax.set_xlabel(\"1st eigenvector\")\n",
    "ax.w_xaxis.set_ticklabels([])\n",
    "ax.set_ylabel(\"2nd eigenvector\")\n",
    "ax.w_yaxis.set_ticklabels([])\n",
    "ax.set_zlabel(\"3rd eigenvector\")\n",
    "ax.w_zaxis.set_ticklabels([])\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 類別型變數onehot encode"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. OneHotEncoding\n",
    "```\n",
    "# from sklearn.preprocessing import OneHotEncoder\n",
    "# [0,1,2] => \n",
    "# [\n",
    "#     [1,0,0],\n",
    "#     [0,1,0],\n",
    "#     [0,0,1],\n",
    "# ]\n",
    "```\n",
    "2. sklearn encoder or classifier usage\n",
    "    1. fit\n",
    "    2. fit_transform\n",
    "    3. transform (or predict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "OneHotEncoder(categorical_features=None, categories=None, drop=None,\n",
       "              dtype=<class 'numpy.float64'>, handle_unknown='error',\n",
       "              n_values=None, sparse=True)"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 請使用preprocessing.OneHotEncoder()初始化一個encoder\n",
    "#=============your works starts===============#\n",
    "enc = preprocessing.OneHotEncoder()\n",
    "#==============your works ends================#\n",
    "\n",
    "enc\n",
    "# OneHotEncoder(categorical_features='all', dtype=<class 'numpy.float64'>,\n",
    "#        handle_unknown='error', n_values='auto', sparse=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1 [1. 0. 0. 0. 0.]\n",
      "3 [0. 0. 1. 0. 0.]\n",
      "2 [0. 1. 0. 0. 0.]\n",
      "3 [0. 0. 1. 0. 0.]\n",
      "1 [1. 0. 0. 0. 0.]\n",
      "1 [1. 0. 0. 0. 0.]\n",
      "1 [1. 0. 0. 0. 0.]\n",
      "4 [0. 0. 0. 1. 0.]\n",
      "3 [0. 0. 1. 0. 0.]\n",
      "3 [0. 0. 1. 0. 0.]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\users\\thinktron\\appdata\\local\\programs\\python\\python36\\lib\\site-packages\\sklearn\\preprocessing\\_encoders.py:415: FutureWarning: The handling of integer data will change in version 0.22. Currently, the categories are determined based on the range [0, max(values)], while in the future they will be determined based on the unique values.\n",
      "If you want the future behaviour and silence this warning, you can specify \"categories='auto'\".\n",
      "In case you used a LabelEncoder before this OneHotEncoder to convert the categories to integers, then you can now use the OneHotEncoder directly.\n",
      "  warnings.warn(msg, FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "# 請使用enc.fit_transform兩個步驟，onehot encode title_cat\n",
    "#=============your works starts===============#\n",
    "title_cat_onehot = enc.fit_transform(df[['title_cat']].values).toarray()\n",
    "#==============your works ends================#\n",
    "\n",
    "for i in range(10):\n",
    "    print(df['title_cat'].values[i], title_cat_onehot[i])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\users\\thinktron\\appdata\\local\\programs\\python\\python36\\lib\\site-packages\\sklearn\\preprocessing\\_encoders.py:415: FutureWarning: The handling of integer data will change in version 0.22. Currently, the categories are determined based on the range [0, max(values)], while in the future they will be determined based on the unique values.\n",
      "If you want the future behaviour and silence this warning, you can specify \"categories='auto'\".\n",
      "In case you used a LabelEncoder before this OneHotEncoder to convert the categories to integers, then you can now use the OneHotEncoder directly.\n",
      "  warnings.warn(msg, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[0., 1., 0., 0.],\n",
       "       [0., 0., 1., 0.],\n",
       "       [0., 1., 0., 0.]])"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "enc = preprocessing.OneHotEncoder()\n",
    "# 請使用enc.fit_transform兩個步驟，onehot encode embarked_cat\n",
    "#=============your works starts===============#\n",
    "embarked_cat_onehot = enc.fit_transform(df[['embarked_cat']].values).toarray()\n",
    "#==============your works ends================#\n",
    "\n",
    "embarked_cat_onehot[:3]\n",
    "# array([[0., 0., 1., 0.],\n",
    "#        [0., 1., 0., 0.],\n",
    "#        [0., 0., 1., 0.]])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# PolynomialFeatures"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.86033801 0.74018149]\n",
      " [1.8529878  3.43356378]]\n"
     ]
    }
   ],
   "source": [
    "poly = preprocessing.PolynomialFeatures(degree=2)\n",
    "# 請利用poly.fit_transform製造出fare_log10的0次項、1次項、2次項，並把0次項拿掉\n",
    "#=============your works starts===============#\n",
    "fare_log10_poly = poly.fit_transform(df[['fare_log10']])[:, 1:]\n",
    "#==============your works ends================#\n",
    "\n",
    "print(fare_log10_poly[:2])\n",
    "# [[0.86033801 0.74018149]\n",
    "#  [1.8529878  3.43356378]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[8.60338007e-01 8.31383556e-04 7.40181486e-01 7.15270871e-04\n",
      "  6.91198616e-07]\n",
      " [1.85298780e+00 1.43602614e-03 3.43356378e+00 2.66093892e-03\n",
      "  2.06217108e-06]]\n"
     ]
    }
   ],
   "source": [
    "# 請利用poly.fit_transform製造出'fare_log10', 'random_age'的二項次及其一次交成項\n",
    "#=============your works starts===============#\n",
    "age_fare_ploy = poly.fit_transform(df[['fare_log10', 'avg_age_l1']])[:, 1:]\n",
    "#==============your works ends================#\n",
    "\n",
    "print(age_fare_ploy[:2])\n",
    "# [[8.60338007e-01 8.31383556e-04 7.40181486e-01 7.15270871e-04 6.91198616e-07]\n",
    "#  [1.85298780e+00 1.43602614e-03 3.43356378e+00 2.66093892e-03 2.06217108e-06]]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Preprocessing Conclude"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['PassengerId', 'Survived', 'Pclass', 'Name', 'Sex', 'Age', 'SibSp',\n",
       "       'Parch', 'Ticket', 'Fare', 'Cabin', 'Embarked', 'avg_age', 'median_age',\n",
       "       'random_age', 'avg_age_l1', 'avg_age_l2', 'avg_age_01', 'cabin_cat',\n",
       "       'sex_cat', 'ticket_cat', 'embarked_cat', 'title_cat', 'name_length',\n",
       "       'fare_log10'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(891, 20)\n",
      "(891, 1)\n"
     ]
    }
   ],
   "source": [
    "X = df[['SibSp', 'Parch', 'avg_age_l2', 'sex_cat', 'name_length', 'fare_log10']].values\n",
    "X = np.concatenate([X, title_cat_onehot, embarked_cat_onehot, age_fare_ploy], axis=1)\n",
    "Y = df[['Survived']].values\n",
    "\n",
    "print(X.shape)\n",
    "print(Y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8071748878923767\n"
     ]
    }
   ],
   "source": [
    "from sklearn import linear_model\n",
    "from  sklearn.model_selection import train_test_split\n",
    "\n",
    "\n",
    "X_train, X_test, Y_train, Y_test = train_test_split(X, Y, random_state=1212)\n",
    "\n",
    "reg = linear_model.LinearRegression()\n",
    "reg.fit(X_train, Y_train)\n",
    "predict_prob = reg.predict(X_test)\n",
    "\n",
    "Y_predict = predict_prob > 0.5 \n",
    "Y_test = Y_test == 1\n",
    "acc = np.sum(Y_predict == Y_test)/ len(Y_test)\n",
    "print(\"Accuracy:\", acc)\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "IIIMaterial",
   "language": "python",
   "name": "iiimaterial"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}