20200911_hahow_pandas.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- author: Lee Meng\n",
    "- date: 2020-09-11 16:00\n",
    "- title: Hahow Pandas Workspace\n",
    "- slug: hahow-pandas\n",
    "- tags: 資料科學, Python, pandas\n",
    "- description: \n",
    "- image: nick-hillier-yD5rv8_WzxA-unsplash.jpg\n",
    "- image_credit_url: \n",
    "- status: draft"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "將接近 40 個實用的 pandas 技巧由淺入深地分成 6 大類別：\n",
    "1. [建立 DataFrame](#建立-DataFrame)\n",
    "2. [客製化 DataFrame 顯示設定](#客製化-DataFrame-顯示設定_1)\n",
    "3. [數據清理 & 整理](#數據清理-&-整理_1)\n",
    "4. [取得想要關注的數據](#取得想要關注的數據_1)\n",
    "5. [基本數據處理與轉換](#基本數據處理與轉換_1)\n",
    "6. [簡單匯總 & 分析數據](#簡單匯總-&-分析數據_1)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# [10 minutes to pandas](https://pandas.pydata.org/docs/user_guide/10min.html#min)\n",
    "- https://pandas.pydata.org/docs/user_guide/10min.html#min"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'1.1.2'"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "pd.__version__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Object creation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    1.0\n",
       "1    3.0\n",
       "2    5.0\n",
       "3    NaN\n",
       "4    6.0\n",
       "5    8.0\n",
       "dtype: float64"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s = pd.Series([1, 3, 5, np.nan, 6, 8])\n",
    "s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DatetimeIndex(['2013-01-01', '2013-01-02', '2013-01-03', '2013-01-04',\n",
       "               '2013-01-05', '2013-01-06'],\n",
       "              dtype='datetime64[ns]', freq='D')"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "date = pd.date_range('20130101', periods=6)\n",
    "date"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.960916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>0.101864</td>\n",
       "      <td>-0.127313</td>\n",
       "      <td>-0.540541</td>\n",
       "      <td>0.338675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.745846</td>\n",
       "      <td>0.182203</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>1.006620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>-0.227424</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>0.347778</td>\n",
       "      <td>1.458274</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>-0.442323</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107\n",
       "2013-01-02  0.949909  2.004353  1.160331 -0.960916\n",
       "2013-01-03  0.101864 -0.127313 -0.540541  0.338675\n",
       "2013-01-04  0.745846  0.182203 -2.515645  1.006620\n",
       "2013-01-05 -0.227424  1.019979  0.347778  1.458274\n",
       "2013-01-06  1.653608 -2.246206  1.150773 -0.442323"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame(np.random.randn(6, 4), index=date, columns=list('ABCD'))\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>E</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2020-09-13</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2</td>\n",
       "      <td>train</td>\n",
       "      <td>Yo</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2020-09-13</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2</td>\n",
       "      <td>test</td>\n",
       "      <td>Yo</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2020-09-13</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2</td>\n",
       "      <td>train</td>\n",
       "      <td>Yo</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2020-09-13</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2</td>\n",
       "      <td>test</td>\n",
       "      <td>Yo</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     A          B    C  D      E   F\n",
       "0  1.0 2020-09-13  5.0  2  train  Yo\n",
       "1  1.0 2020-09-13  5.0  2   test  Yo\n",
       "2  1.0 2020-09-13  5.0  2  train  Yo\n",
       "3  1.0 2020-09-13  5.0  2   test  Yo"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2 = pd.DataFrame({\n",
    "    'A': 1.,\n",
    "    'B': pd.Timestamp('20200913'),\n",
    "    'C': pd.Series(5, index=list(range(4)), dtype='float32'),\n",
    "    'D': np.array([2] * 4, dtype='int32'),\n",
    "    'E': pd.Categorical(['train', 'test'] * 2),\n",
    "    'F': 'Yo'\n",
    "})\n",
    "df2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "A           float64\n",
       "B    datetime64[ns]\n",
       "C           float32\n",
       "D             int32\n",
       "E          category\n",
       "F            object\n",
       "dtype: object"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Viewing Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>E</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2020-09-13</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2</td>\n",
       "      <td>train</td>\n",
       "      <td>Yo</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2020-09-13</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2</td>\n",
       "      <td>test</td>\n",
       "      <td>Yo</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2020-09-13</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2</td>\n",
       "      <td>train</td>\n",
       "      <td>Yo</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2020-09-13</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2</td>\n",
       "      <td>test</td>\n",
       "      <td>Yo</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     A          B    C  D      E   F\n",
       "0  1.0 2020-09-13  5.0  2  train  Yo\n",
       "1  1.0 2020-09-13  5.0  2   test  Yo\n",
       "2  1.0 2020-09-13  5.0  2  train  Yo\n",
       "3  1.0 2020-09-13  5.0  2   test  Yo"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.head() # df.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Int64Index([0, 1, 2, 3], dtype='int64')"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.index # df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DatetimeIndex(['2020-09-13', '2020-09-14', '2020-09-15', '2020-09-16'], dtype='datetime64[ns]', freq='D')"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.index = pd.date_range('20200913', periods=df2.index.size)\n",
    "df2.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 1.60701278, -0.60619109, -1.79429157, -0.35310702],\n",
       "       [ 0.94990932,  2.00435303,  1.1603309 , -0.96091587],\n",
       "       [ 0.10186365, -0.12731272, -0.54054077,  0.33867545],\n",
       "       [ 0.74584567,  0.18220302, -2.51564452,  1.00661959],\n",
       "       [-0.22742367,  1.01997864,  0.34777848,  1.45827367],\n",
       "       [ 1.65360757, -2.24620584,  1.15077324, -0.44232284]])"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.to_numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>6.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>6.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.805136</td>\n",
       "      <td>0.037804</td>\n",
       "      <td>-0.365266</td>\n",
       "      <td>0.174537</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.767875</td>\n",
       "      <td>1.449933</td>\n",
       "      <td>1.538260</td>\n",
       "      <td>0.929027</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-0.227424</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>-0.960916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.262859</td>\n",
       "      <td>-0.486472</td>\n",
       "      <td>-1.480854</td>\n",
       "      <td>-0.420019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.847877</td>\n",
       "      <td>0.027445</td>\n",
       "      <td>-0.096381</td>\n",
       "      <td>-0.007216</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.442737</td>\n",
       "      <td>0.810535</td>\n",
       "      <td>0.950025</td>\n",
       "      <td>0.839634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>1.458274</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              A         B         C         D\n",
       "count  6.000000  6.000000  6.000000  6.000000\n",
       "mean   0.805136  0.037804 -0.365266  0.174537\n",
       "std    0.767875  1.449933  1.538260  0.929027\n",
       "min   -0.227424 -2.246206 -2.515645 -0.960916\n",
       "25%    0.262859 -0.486472 -1.480854 -0.420019\n",
       "50%    0.847877  0.027445 -0.096381 -0.007216\n",
       "75%    1.442737  0.810535  0.950025  0.839634\n",
       "max    1.653608  2.004353  1.160331  1.458274"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "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>2013-01-01</th>\n",
       "      <th>2013-01-02</th>\n",
       "      <th>2013-01-03</th>\n",
       "      <th>2013-01-04</th>\n",
       "      <th>2013-01-05</th>\n",
       "      <th>2013-01-06</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>A</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>0.949909</td>\n",
       "      <td>0.101864</td>\n",
       "      <td>0.745846</td>\n",
       "      <td>-0.227424</td>\n",
       "      <td>1.653608</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>B</th>\n",
       "      <td>-0.606191</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>-0.127313</td>\n",
       "      <td>0.182203</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>-2.246206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>C</th>\n",
       "      <td>-1.794292</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.540541</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>0.347778</td>\n",
       "      <td>1.150773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>D</th>\n",
       "      <td>-0.353107</td>\n",
       "      <td>-0.960916</td>\n",
       "      <td>0.338675</td>\n",
       "      <td>1.006620</td>\n",
       "      <td>1.458274</td>\n",
       "      <td>-0.442323</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   2013-01-01  2013-01-02  2013-01-03  2013-01-04  2013-01-05  2013-01-06\n",
       "A    1.607013    0.949909    0.101864    0.745846   -0.227424    1.653608\n",
       "B   -0.606191    2.004353   -0.127313    0.182203    1.019979   -2.246206\n",
       "C   -1.794292    1.160331   -0.540541   -2.515645    0.347778    1.150773\n",
       "D   -0.353107   -0.960916    0.338675    1.006620    1.458274   -0.442323"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "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>D</th>\n",
       "      <th>C</th>\n",
       "      <th>B</th>\n",
       "      <th>A</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>-0.353107</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>1.607013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>-0.960916</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>0.949909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>0.338675</td>\n",
       "      <td>-0.540541</td>\n",
       "      <td>-0.127313</td>\n",
       "      <td>0.101864</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>1.006620</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>0.182203</td>\n",
       "      <td>0.745846</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>1.458274</td>\n",
       "      <td>0.347778</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>-0.227424</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>-0.442323</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>1.653608</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   D         C         B         A\n",
       "2013-01-01 -0.353107 -1.794292 -0.606191  1.607013\n",
       "2013-01-02 -0.960916  1.160331  2.004353  0.949909\n",
       "2013-01-03  0.338675 -0.540541 -0.127313  0.101864\n",
       "2013-01-04  1.006620 -2.515645  0.182203  0.745846\n",
       "2013-01-05  1.458274  0.347778  1.019979 -0.227424\n",
       "2013-01-06 -0.442323  1.150773 -2.246206  1.653608"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sort_index(axis=1, ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.960916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>-0.442323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>0.101864</td>\n",
       "      <td>-0.127313</td>\n",
       "      <td>-0.540541</td>\n",
       "      <td>0.338675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.745846</td>\n",
       "      <td>0.182203</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>1.006620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>-0.227424</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>0.347778</td>\n",
       "      <td>1.458274</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "2013-01-02  0.949909  2.004353  1.160331 -0.960916\n",
       "2013-01-06  1.653608 -2.246206  1.150773 -0.442323\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107\n",
       "2013-01-03  0.101864 -0.127313 -0.540541  0.338675\n",
       "2013-01-04  0.745846  0.182203 -2.515645  1.006620\n",
       "2013-01-05 -0.227424  1.019979  0.347778  1.458274"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sort_values(by='D')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## Selection"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "- production 時推薦使用 .loc, .iloc 等 methods"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Getting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2013-01-01    1.607013\n",
       "2013-01-02    0.949909\n",
       "2013-01-03    0.101864\n",
       "2013-01-04    0.745846\n",
       "2013-01-05   -0.227424\n",
       "2013-01-06    1.653608\n",
       "Freq: D, Name: A, dtype: float64"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['A'] # df.A"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [A, B, C, D]\n",
       "Index: []"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['2020-09-14':'2020-09-15']  #row-wise"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Selection by Label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timestamp('2013-01-01 00:00:00', freq='D')"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "date[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.960916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>0.101864</td>\n",
       "      <td>-0.127313</td>\n",
       "      <td>-0.540541</td>\n",
       "      <td>0.338675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.745846</td>\n",
       "      <td>0.182203</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>1.006620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>-0.227424</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>0.347778</td>\n",
       "      <td>1.458274</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>-0.442323</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107\n",
       "2013-01-02  0.949909  2.004353  1.160331 -0.960916\n",
       "2013-01-03  0.101864 -0.127313 -0.540541  0.338675\n",
       "2013-01-04  0.745846  0.182203 -2.515645  1.006620\n",
       "2013-01-05 -0.227424  1.019979  0.347778  1.458274\n",
       "2013-01-06  1.653608 -2.246206  1.150773 -0.442323"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "A    1.607013\n",
       "B   -0.606191\n",
       "C   -1.794292\n",
       "D   -0.353107\n",
       "Name: 2013-01-01 00:00:00, dtype: float64"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.loc[date[0]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>0.101864</td>\n",
       "      <td>-0.127313</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.745846</td>\n",
       "      <td>0.182203</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>-0.227424</td>\n",
       "      <td>1.019979</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>-2.246206</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B\n",
       "2013-01-01  1.607013 -0.606191\n",
       "2013-01-02  0.949909  2.004353\n",
       "2013-01-03  0.101864 -0.127313\n",
       "2013-01-04  0.745846  0.182203\n",
       "2013-01-05 -0.227424  1.019979\n",
       "2013-01-06  1.653608 -2.246206"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.loc[:, ['A', 'B']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "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>B</th>\n",
       "      <th>C</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>-0.127313</td>\n",
       "      <td>-0.540541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.182203</td>\n",
       "      <td>-2.515645</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   B         C\n",
       "2013-01-03 -0.127313 -0.540541\n",
       "2013-01-04  0.182203 -2.515645"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.loc['2013-01-03':'2013-01-04', ['B', 'C']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "12.2 µs ± 290 ns per loop (mean ± std. dev. of 7 runs, 100000 loops each)\n"
     ]
    }
   ],
   "source": [
    "%timeit df.at[date[0], 'B']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "14.9 µs ± 930 ns per loop (mean ± std. dev. of 7 runs, 10000 loops each)\n"
     ]
    }
   ],
   "source": [
    "%timeit df.loc[date[0], 'B']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Select by Position"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.960916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>0.101864</td>\n",
       "      <td>-0.127313</td>\n",
       "      <td>-0.540541</td>\n",
       "      <td>0.338675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.745846</td>\n",
       "      <td>0.182203</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>1.006620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>-0.227424</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>0.347778</td>\n",
       "      <td>1.458274</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>-0.442323</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107\n",
       "2013-01-02  0.949909  2.004353  1.160331 -0.960916\n",
       "2013-01-03  0.101864 -0.127313 -0.540541  0.338675\n",
       "2013-01-04  0.745846  0.182203 -2.515645  1.006620\n",
       "2013-01-05 -0.227424  1.019979  0.347778  1.458274\n",
       "2013-01-06  1.653608 -2.246206  1.150773 -0.442323"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.960916</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107\n",
       "2013-01-02  0.949909  2.004353  1.160331 -0.960916"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.iloc[:2]  # row"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "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>B</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-0.353107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>2.004353</td>\n",
       "      <td>-0.960916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>-0.127313</td>\n",
       "      <td>0.338675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.182203</td>\n",
       "      <td>1.006620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>1.019979</td>\n",
       "      <td>1.458274</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>-2.246206</td>\n",
       "      <td>-0.442323</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   B         D\n",
       "2013-01-01 -0.606191 -0.353107\n",
       "2013-01-02  2.004353 -0.960916\n",
       "2013-01-03 -0.127313  0.338675\n",
       "2013-01-04  0.182203  1.006620\n",
       "2013-01-05  1.019979  1.458274\n",
       "2013-01-06 -2.246206 -0.442323"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.iloc[:, [1, 3]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "15.8 µs ± 863 ns per loop (mean ± std. dev. of 7 runs, 10000 loops each)\n"
     ]
    }
   ],
   "source": [
    "%timeit df.iloc[1, 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "14 µs ± 869 ns per loop (mean ± std. dev. of 7 runs, 10000 loops each)\n"
     ]
    }
   ],
   "source": [
    "%timeit df.iat[1, 1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Boolean Indexing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>-0.442323</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107\n",
       "2013-01-06  1.653608 -2.246206  1.150773 -0.442323"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df['A'] > 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>NaN</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.006620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.458274</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "2013-01-01  1.607013       NaN       NaN       NaN\n",
       "2013-01-02       NaN  2.004353  1.160331       NaN\n",
       "2013-01-03       NaN       NaN       NaN       NaN\n",
       "2013-01-04       NaN       NaN       NaN  1.006620\n",
       "2013-01-05       NaN  1.019979       NaN  1.458274\n",
       "2013-01-06  1.653608       NaN  1.150773       NaN"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df > 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>E</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2020-09-13</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2020-09-13</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2</td>\n",
       "      <td>train</td>\n",
       "      <td>Yo</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-09-15</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2020-09-13</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2</td>\n",
       "      <td>train</td>\n",
       "      <td>Yo</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              A          B    C  D      E   F\n",
       "2020-09-13  1.0 2020-09-13  5.0  2  train  Yo\n",
       "2020-09-15  1.0 2020-09-13  5.0  2  train  Yo"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2[df2['E'].isin(['train'])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>E</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2020-09-13</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2020-09-13</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2</td>\n",
       "      <td>train</td>\n",
       "      <td>Yo</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-09-15</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2020-09-13</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2</td>\n",
       "      <td>train</td>\n",
       "      <td>Yo</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              A          B    C  D      E   F\n",
       "2020-09-13  1.0 2020-09-13  5.0  2  train  Yo\n",
       "2020-09-15  1.0 2020-09-13  5.0  2  train  Yo"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2[df2['E'] == 'train']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Setting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.960916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>0.101864</td>\n",
       "      <td>-0.127313</td>\n",
       "      <td>-0.540541</td>\n",
       "      <td>0.338675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.745846</td>\n",
       "      <td>0.182203</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>1.006620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>-0.227424</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>0.347778</td>\n",
       "      <td>1.458274</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>-0.442323</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107\n",
       "2013-01-02  0.949909  2.004353  1.160331 -0.960916\n",
       "2013-01-03  0.101864 -0.127313 -0.540541  0.338675\n",
       "2013-01-04  0.745846  0.182203 -2.515645  1.006620\n",
       "2013-01-05 -0.227424  1.019979  0.347778  1.458274\n",
       "2013-01-06  1.653608 -2.246206  1.150773 -0.442323"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2013-01-02    2\n",
       "2013-01-03    3\n",
       "2013-01-04    4\n",
       "2013-01-05    5\n",
       "2013-01-06    6\n",
       "2013-01-07    7\n",
       "Freq: D, dtype: int64"
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s1 = pd.Series(range(2, 8), index=pd.date_range('2013-01-02', periods=6))\n",
    "s1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.960916</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>0.101864</td>\n",
       "      <td>-0.127313</td>\n",
       "      <td>-0.540541</td>\n",
       "      <td>0.338675</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.745846</td>\n",
       "      <td>0.182203</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>1.006620</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>-0.227424</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>0.347778</td>\n",
       "      <td>1.458274</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>-0.442323</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D    F\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107  NaN\n",
       "2013-01-02  0.949909  2.004353  1.160331 -0.960916  2.0\n",
       "2013-01-03  0.101864 -0.127313 -0.540541  0.338675  3.0\n",
       "2013-01-04  0.745846  0.182203 -2.515645  1.006620  4.0\n",
       "2013-01-05 -0.227424  1.019979  0.347778  1.458274  5.0\n",
       "2013-01-06  1.653608 -2.246206  1.150773 -0.442323  6.0"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['F'] = s1\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.960916</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.127313</td>\n",
       "      <td>-0.540541</td>\n",
       "      <td>0.338675</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.745846</td>\n",
       "      <td>0.182203</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>1.006620</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>-0.227424</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>0.347778</td>\n",
       "      <td>1.458274</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>-0.442323</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D    F\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107  NaN\n",
       "2013-01-02  0.949909  2.004353  1.160331 -0.960916  2.0\n",
       "2013-01-03  0.000000 -0.127313 -0.540541  0.338675  3.0\n",
       "2013-01-04  0.745846  0.182203 -2.515645  1.006620  4.0\n",
       "2013-01-05 -0.227424  1.019979  0.347778  1.458274  5.0\n",
       "2013-01-06  1.653608 -2.246206  1.150773 -0.442323  6.0"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.at['2013-01-03', 'A'] = 0\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.960916</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.127313</td>\n",
       "      <td>-0.540541</td>\n",
       "      <td>0.338675</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.745846</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>1.006620</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>-0.227424</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>0.347778</td>\n",
       "      <td>1.458274</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>-0.442323</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D    F\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107  NaN\n",
       "2013-01-02  0.949909  2.004353  1.160331 -0.960916  2.0\n",
       "2013-01-03  0.000000 -0.127313 -0.540541  0.338675  3.0\n",
       "2013-01-04  0.745846  0.000000 -2.515645  1.006620  4.0\n",
       "2013-01-05 -0.227424  1.019979  0.347778  1.458274  5.0\n",
       "2013-01-06  1.653608 -2.246206  1.150773 -0.442323  6.0"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.iat[3, 1] = 0\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.338675</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.745846</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.006620</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>0.347778</td>\n",
       "      <td>1.458274</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D    F\n",
       "2013-01-01  1.607013       NaN       NaN       NaN  NaN\n",
       "2013-01-02  0.949909  2.004353  1.160331       NaN  2.0\n",
       "2013-01-03       NaN       NaN       NaN  0.338675  3.0\n",
       "2013-01-04  0.745846       NaN       NaN  1.006620  4.0\n",
       "2013-01-05       NaN  1.019979  0.347778  1.458274  5.0\n",
       "2013-01-06  1.653608       NaN  1.150773       NaN  6.0"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2 = df.copy()\n",
    "df2[df2 > 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1000.000000</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>-0.960916</td>\n",
       "      <td>1000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.127313</td>\n",
       "      <td>-0.540541</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>1000.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>-0.227424</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1000.000000</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>-0.442323</td>\n",
       "      <td>1000.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      A            B            C            D       F\n",
       "2013-01-01  1000.000000    -0.606191    -1.794292    -0.353107     NaN\n",
       "2013-01-02  1000.000000  1000.000000  1000.000000    -0.960916  1000.0\n",
       "2013-01-03     0.000000    -0.127313    -0.540541  1000.000000  1000.0\n",
       "2013-01-04  1000.000000     0.000000    -2.515645  1000.000000  1000.0\n",
       "2013-01-05    -0.227424  1000.000000  1000.000000  1000.000000  1000.0\n",
       "2013-01-06  1000.000000    -2.246206  1000.000000    -0.442323  1000.0"
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2[df2 > 0] = 1000\n",
    "df2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## Missing Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.960916</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.127313</td>\n",
       "      <td>-0.540541</td>\n",
       "      <td>0.338675</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.745846</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>1.006620</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>-0.227424</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>0.347778</td>\n",
       "      <td>1.458274</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>-0.442323</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D    F\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107  NaN\n",
       "2013-01-02  0.949909  2.004353  1.160331 -0.960916  2.0\n",
       "2013-01-03  0.000000 -0.127313 -0.540541  0.338675  3.0\n",
       "2013-01-04  0.745846  0.000000 -2.515645  1.006620  4.0\n",
       "2013-01-05 -0.227424  1.019979  0.347778  1.458274  5.0\n",
       "2013-01-06  1.653608 -2.246206  1.150773 -0.442323  6.0"
      ]
     },
     "execution_count": 120,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>F</th>\n",
       "      <th>New</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "      <td>10</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.960916</td>\n",
       "      <td>10</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D   F  New\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107  10  2.0\n",
       "2013-01-02  0.949909  2.004353  1.160331 -0.960916  10  NaN"
      ]
     },
     "execution_count": 127,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1 = df.reindex(index=date[:2], columns=list(df.columns) + ['New'])\n",
    "df1.loc['2013-01-01', 'New'] = 2\n",
    "df1['F'] = 10\n",
    "df1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>F</th>\n",
       "      <th>New</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "      <td>10</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D   F  New\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107  10  2.0"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1.dropna(how='any')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>F</th>\n",
       "      <th>New</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.960916</td>\n",
       "      <td>10</td>\n",
       "      <td>1234</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D   F   New\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107  10     2\n",
       "2013-01-02  0.949909  2.004353  1.160331 -0.960916  10  1234"
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1.fillna(value='1234')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>F</th>\n",
       "      <th>New</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                A      B      C      D      F    New\n",
       "2013-01-01  False  False  False  False  False  False\n",
       "2013-01-02  False  False  False  False  False   True"
      ]
     },
     "execution_count": 130,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.isna(df1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## Operations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.960916</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.127313</td>\n",
       "      <td>-0.540541</td>\n",
       "      <td>0.338675</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.745846</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>1.006620</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>-0.227424</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>0.347778</td>\n",
       "      <td>1.458274</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>-0.442323</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D    F\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107  NaN\n",
       "2013-01-02  0.949909  2.004353  1.160331 -0.960916  2.0\n",
       "2013-01-03  0.000000 -0.127313 -0.540541  0.338675  3.0\n",
       "2013-01-04  0.745846  0.000000 -2.515645  1.006620  4.0\n",
       "2013-01-05 -0.227424  1.019979  0.347778  1.458274  5.0\n",
       "2013-01-06  1.653608 -2.246206  1.150773 -0.442323  6.0"
      ]
     },
     "execution_count": 132,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "A    0.788159\n",
       "B    0.007437\n",
       "C   -0.365266\n",
       "D    0.174537\n",
       "F    4.000000\n",
       "dtype: float64"
      ]
     },
     "execution_count": 133,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2013-01-01   -0.286644\n",
       "2013-01-02    1.030735\n",
       "2013-01-03    0.534164\n",
       "2013-01-04    0.647364\n",
       "2013-01-05    1.519721\n",
       "2013-01-06    1.223170\n",
       "Freq: D, dtype: float64"
      ]
     },
     "execution_count": 134,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.mean(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.960916</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.127313</td>\n",
       "      <td>-0.540541</td>\n",
       "      <td>0.338675</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.745846</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>1.006620</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>-0.227424</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>0.347778</td>\n",
       "      <td>1.458274</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>-0.442323</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D    F\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107  NaN\n",
       "2013-01-02  0.949909  2.004353  1.160331 -0.960916  2.0\n",
       "2013-01-03  0.000000 -0.127313 -0.540541  0.338675  3.0\n",
       "2013-01-04  0.745846  0.000000 -2.515645  1.006620  4.0\n",
       "2013-01-05 -0.227424  1.019979  0.347778  1.458274  5.0\n",
       "2013-01-06  1.653608 -2.246206  1.150773 -0.442323  6.0"
      ]
     },
     "execution_count": 135,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2013-01-01    NaN\n",
       "2013-01-02    NaN\n",
       "2013-01-03    1.0\n",
       "2013-01-04    3.0\n",
       "2013-01-05    5.0\n",
       "2013-01-06    NaN\n",
       "Freq: D, dtype: float64"
      ]
     },
     "execution_count": 137,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s = pd.Series([1, 3, 5, np.nan, 6, 8], index=date).shift(2)\n",
    "s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>-1.000000</td>\n",
       "      <td>-1.127313</td>\n",
       "      <td>-1.540541</td>\n",
       "      <td>-0.661325</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>-2.254154</td>\n",
       "      <td>-3.000000</td>\n",
       "      <td>-5.515645</td>\n",
       "      <td>-1.993380</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>-5.227424</td>\n",
       "      <td>-3.980021</td>\n",
       "      <td>-4.652222</td>\n",
       "      <td>-3.541726</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D    F\n",
       "2013-01-01       NaN       NaN       NaN       NaN  NaN\n",
       "2013-01-02       NaN       NaN       NaN       NaN  NaN\n",
       "2013-01-03 -1.000000 -1.127313 -1.540541 -0.661325  2.0\n",
       "2013-01-04 -2.254154 -3.000000 -5.515645 -1.993380  1.0\n",
       "2013-01-05 -5.227424 -3.980021 -4.652222 -3.541726  0.0\n",
       "2013-01-06       NaN       NaN       NaN       NaN  NaN"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sub(s, axis='index')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Apply"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>0.949909</td>\n",
       "      <td>2.004353</td>\n",
       "      <td>1.160331</td>\n",
       "      <td>-0.960916</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.127313</td>\n",
       "      <td>-0.540541</td>\n",
       "      <td>0.338675</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>0.745846</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-2.515645</td>\n",
       "      <td>1.006620</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>-0.227424</td>\n",
       "      <td>1.019979</td>\n",
       "      <td>0.347778</td>\n",
       "      <td>1.458274</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>1.653608</td>\n",
       "      <td>-2.246206</td>\n",
       "      <td>1.150773</td>\n",
       "      <td>-0.442323</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D    F\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107  NaN\n",
       "2013-01-02  0.949909  2.004353  1.160331 -0.960916  2.0\n",
       "2013-01-03  0.000000 -0.127313 -0.540541  0.338675  3.0\n",
       "2013-01-04  0.745846  0.000000 -2.515645  1.006620  4.0\n",
       "2013-01-05 -0.227424  1.019979  0.347778  1.458274  5.0\n",
       "2013-01-06  1.653608 -2.246206  1.150773 -0.442323  6.0"
      ]
     },
     "execution_count": 140,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-01-01</th>\n",
       "      <td>1.607013</td>\n",
       "      <td>-0.606191</td>\n",
       "      <td>-1.794292</td>\n",
       "      <td>-0.353107</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-02</th>\n",
       "      <td>2.556922</td>\n",
       "      <td>1.398162</td>\n",
       "      <td>-0.633961</td>\n",
       "      <td>-1.314023</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-03</th>\n",
       "      <td>2.556922</td>\n",
       "      <td>1.270849</td>\n",
       "      <td>-1.174501</td>\n",
       "      <td>-0.975347</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-04</th>\n",
       "      <td>3.302768</td>\n",
       "      <td>1.270849</td>\n",
       "      <td>-3.690146</td>\n",
       "      <td>0.031272</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-05</th>\n",
       "      <td>3.075344</td>\n",
       "      <td>2.290828</td>\n",
       "      <td>-3.342367</td>\n",
       "      <td>1.489546</td>\n",
       "      <td>14.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-06</th>\n",
       "      <td>4.728952</td>\n",
       "      <td>0.044622</td>\n",
       "      <td>-2.191594</td>\n",
       "      <td>1.047223</td>\n",
       "      <td>20.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D     F\n",
       "2013-01-01  1.607013 -0.606191 -1.794292 -0.353107   NaN\n",
       "2013-01-02  2.556922  1.398162 -0.633961 -1.314023   2.0\n",
       "2013-01-03  2.556922  1.270849 -1.174501 -0.975347   5.0\n",
       "2013-01-04  3.302768  1.270849 -3.690146  0.031272   9.0\n",
       "2013-01-05  3.075344  2.290828 -3.342367  1.489546  14.0\n",
       "2013-01-06  4.728952  0.044622 -2.191594  1.047223  20.0"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.apply(np.cumsum) # column wise, axis = index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2013-01-01    3.401304\n",
       "2013-01-02    2.965269\n",
       "2013-01-03    3.540541\n",
       "2013-01-04    6.515645\n",
       "2013-01-05    5.227424\n",
       "2013-01-06    8.246206\n",
       "Freq: D, dtype: float64"
      ]
     },
     "execution_count": 144,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.apply(lambda x: x.max() - x.min(), axis=1) # axis = columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Histogramming"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "62    2\n",
       "44    2\n",
       "31    1\n",
       "57    1\n",
       "69    1\n",
       "68    1\n",
       "60    1\n",
       "64    1\n",
       "dtype: int64"
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s = pd.Series(np.random.randint(10, 70, size=10))\n",
    "s.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### String Methods"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       A\n",
       "1       B\n",
       "2       C\n",
       "3    Aaba\n",
       "4    Baca\n",
       "5     NaN\n",
       "6    CABA\n",
       "7     dog\n",
       "8     cat\n",
       "dtype: object"
      ]
     },
     "execution_count": 150,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s = pd.Series(['A', 'B', 'C', 'Aaba', 'Baca', np.nan, 'CABA', 'dog', 'cat'])\n",
    "s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       a\n",
       "1       b\n",
       "2       c\n",
       "3    aaba\n",
       "4    baca\n",
       "5     NaN\n",
       "6    caba\n",
       "7     dog\n",
       "8     cat\n",
       "dtype: object"
      ]
     },
     "execution_count": 151,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s.str.lower()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Merge"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Concat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.208480</td>\n",
       "      <td>-0.399535</td>\n",
       "      <td>0.170227</td>\n",
       "      <td>1.582362</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.571295</td>\n",
       "      <td>0.824710</td>\n",
       "      <td>0.395026</td>\n",
       "      <td>0.190229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.377516</td>\n",
       "      <td>1.574576</td>\n",
       "      <td>-1.359111</td>\n",
       "      <td>-1.376349</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.145082</td>\n",
       "      <td>-1.783859</td>\n",
       "      <td>0.394369</td>\n",
       "      <td>-0.666709</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.257159</td>\n",
       "      <td>1.208557</td>\n",
       "      <td>1.891070</td>\n",
       "      <td>-0.624411</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>-0.271946</td>\n",
       "      <td>1.068839</td>\n",
       "      <td>-0.300622</td>\n",
       "      <td>0.124577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.226121</td>\n",
       "      <td>0.379922</td>\n",
       "      <td>-1.985552</td>\n",
       "      <td>0.700439</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.732998</td>\n",
       "      <td>0.222775</td>\n",
       "      <td>-0.858922</td>\n",
       "      <td>-1.222658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-1.663526</td>\n",
       "      <td>-0.408940</td>\n",
       "      <td>0.503871</td>\n",
       "      <td>0.498664</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.708381</td>\n",
       "      <td>-0.171467</td>\n",
       "      <td>-0.835987</td>\n",
       "      <td>0.081464</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2         3\n",
       "0  1.208480 -0.399535  0.170227  1.582362\n",
       "1  0.571295  0.824710  0.395026  0.190229\n",
       "2 -1.377516  1.574576 -1.359111 -1.376349\n",
       "3 -0.145082 -1.783859  0.394369 -0.666709\n",
       "4  0.257159  1.208557  1.891070 -0.624411\n",
       "5 -0.271946  1.068839 -0.300622  0.124577\n",
       "6  0.226121  0.379922 -1.985552  0.700439\n",
       "7  1.732998  0.222775 -0.858922 -1.222658\n",
       "8 -1.663526 -0.408940  0.503871  0.498664\n",
       "9  1.708381 -0.171467 -0.835987  0.081464"
      ]
     },
     "execution_count": 152,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame(np.random.randn(10, 4))\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.208480</td>\n",
       "      <td>-0.399535</td>\n",
       "      <td>0.170227</td>\n",
       "      <td>1.582362</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.571295</td>\n",
       "      <td>0.824710</td>\n",
       "      <td>0.395026</td>\n",
       "      <td>0.190229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.377516</td>\n",
       "      <td>1.574576</td>\n",
       "      <td>-1.359111</td>\n",
       "      <td>-1.376349</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2         3\n",
       "0  1.208480 -0.399535  0.170227  1.582362\n",
       "1  0.571295  0.824710  0.395026  0.190229\n",
       "2 -1.377516  1.574576 -1.359111 -1.376349"
      ]
     },
     "execution_count": 153,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pieces = [df[:3], df[3:7], df[7:]]\n",
    "pieces[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.208480</td>\n",
       "      <td>-0.399535</td>\n",
       "      <td>0.170227</td>\n",
       "      <td>1.582362</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.571295</td>\n",
       "      <td>0.824710</td>\n",
       "      <td>0.395026</td>\n",
       "      <td>0.190229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1.377516</td>\n",
       "      <td>1.574576</td>\n",
       "      <td>-1.359111</td>\n",
       "      <td>-1.376349</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.145082</td>\n",
       "      <td>-1.783859</td>\n",
       "      <td>0.394369</td>\n",
       "      <td>-0.666709</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.257159</td>\n",
       "      <td>1.208557</td>\n",
       "      <td>1.891070</td>\n",
       "      <td>-0.624411</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>-0.271946</td>\n",
       "      <td>1.068839</td>\n",
       "      <td>-0.300622</td>\n",
       "      <td>0.124577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.226121</td>\n",
       "      <td>0.379922</td>\n",
       "      <td>-1.985552</td>\n",
       "      <td>0.700439</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.732998</td>\n",
       "      <td>0.222775</td>\n",
       "      <td>-0.858922</td>\n",
       "      <td>-1.222658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-1.663526</td>\n",
       "      <td>-0.408940</td>\n",
       "      <td>0.503871</td>\n",
       "      <td>0.498664</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.708381</td>\n",
       "      <td>-0.171467</td>\n",
       "      <td>-0.835987</td>\n",
       "      <td>0.081464</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2         3\n",
       "0  1.208480 -0.399535  0.170227  1.582362\n",
       "1  0.571295  0.824710  0.395026  0.190229\n",
       "2 -1.377516  1.574576 -1.359111 -1.376349\n",
       "3 -0.145082 -1.783859  0.394369 -0.666709\n",
       "4  0.257159  1.208557  1.891070 -0.624411\n",
       "5 -0.271946  1.068839 -0.300622  0.124577\n",
       "6  0.226121  0.379922 -1.985552  0.700439\n",
       "7  1.732998  0.222775 -0.858922 -1.222658\n",
       "8 -1.663526 -0.408940  0.503871  0.498664\n",
       "9  1.708381 -0.171467 -0.835987  0.081464"
      ]
     },
     "execution_count": 154,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.concat(pieces)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Join"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "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>key</th>\n",
       "      <th>lval</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>foo</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>foo</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   key  lval\n",
       "0  foo     1\n",
       "1  foo     2"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>key</th>\n",
       "      <th>rval</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>foo</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>foo</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   key  rval\n",
       "0  foo     4\n",
       "1  foo     5"
      ]
     },
     "execution_count": 155,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import display\n",
    "\n",
    "left = pd.DataFrame({'key': ['foo', 'foo'], 'lval': [1, 2]})\n",
    "right = pd.DataFrame({'key': ['foo', 'foo'], 'rval': [4, 5]})\n",
    "\n",
    "display(left)\n",
    "right"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "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>key</th>\n",
       "      <th>lval</th>\n",
       "      <th>rval</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>foo</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>foo</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>foo</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>foo</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   key  lval  rval\n",
       "0  foo     1     4\n",
       "1  foo     1     5\n",
       "2  foo     2     4\n",
       "3  foo     2     5"
      ]
     },
     "execution_count": 156,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.merge(left, right, how='left')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## Grouping"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- splitting -> applying -> combining"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>foo</td>\n",
       "      <td>one</td>\n",
       "      <td>0.203450</td>\n",
       "      <td>-0.110570</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>bar</td>\n",
       "      <td>one</td>\n",
       "      <td>0.196015</td>\n",
       "      <td>0.153670</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>foo</td>\n",
       "      <td>two</td>\n",
       "      <td>-1.941297</td>\n",
       "      <td>-1.765952</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>bar</td>\n",
       "      <td>three</td>\n",
       "      <td>-0.253213</td>\n",
       "      <td>-0.895575</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>foo</td>\n",
       "      <td>two</td>\n",
       "      <td>-0.744677</td>\n",
       "      <td>0.719799</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>bar</td>\n",
       "      <td>two</td>\n",
       "      <td>0.318505</td>\n",
       "      <td>-1.546736</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>foo</td>\n",
       "      <td>one</td>\n",
       "      <td>0.646812</td>\n",
       "      <td>-0.594465</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>foo</td>\n",
       "      <td>three</td>\n",
       "      <td>-1.050345</td>\n",
       "      <td>-1.188455</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     A      B         C         D\n",
       "0  foo    one  0.203450 -0.110570\n",
       "1  bar    one  0.196015  0.153670\n",
       "2  foo    two -1.941297 -1.765952\n",
       "3  bar  three -0.253213 -0.895575\n",
       "4  foo    two -0.744677  0.719799\n",
       "5  bar    two  0.318505 -1.546736\n",
       "6  foo    one  0.646812 -0.594465\n",
       "7  foo  three -1.050345 -1.188455"
      ]
     },
     "execution_count": 157,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame({'A': ['foo', 'bar', 'foo', 'bar',\n",
    "                         'foo', 'bar', 'foo', 'foo'],\n",
    "                   'B': ['one', 'one', 'two', 'three',\n",
    "                          'two', 'two', 'one', 'three'],\n",
    "                   'C': np.random.randn(8),\n",
    "                   'D': np.random.randn(8)})\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "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>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>bar</th>\n",
       "      <td>0.261307</td>\n",
       "      <td>-2.288641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>foo</th>\n",
       "      <td>-2.886057</td>\n",
       "      <td>-2.939643</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            C         D\n",
       "A                      \n",
       "bar  0.261307 -2.288641\n",
       "foo -2.886057 -2.939643"
      ]
     },
     "execution_count": 158,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby(\"A\").sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "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></th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">bar</th>\n",
       "      <th>one</th>\n",
       "      <td>0.196015</td>\n",
       "      <td>0.153670</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>three</th>\n",
       "      <td>-0.253213</td>\n",
       "      <td>-0.895575</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>two</th>\n",
       "      <td>0.318505</td>\n",
       "      <td>-1.546736</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">foo</th>\n",
       "      <th>one</th>\n",
       "      <td>0.850262</td>\n",
       "      <td>-0.705036</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>three</th>\n",
       "      <td>-1.050345</td>\n",
       "      <td>-1.188455</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>two</th>\n",
       "      <td>-2.685974</td>\n",
       "      <td>-1.046153</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  C         D\n",
       "A   B                        \n",
       "bar one    0.196015  0.153670\n",
       "    three -0.253213 -0.895575\n",
       "    two    0.318505 -1.546736\n",
       "foo one    0.850262 -0.705036\n",
       "    three -1.050345 -1.188455\n",
       "    two   -2.685974 -1.046153"
      ]
     },
     "execution_count": 159,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby(['A', 'B']).sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## Reshaping"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Stack"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('bar', 'one'),\n",
       " ('bar', 'two'),\n",
       " ('baz', 'one'),\n",
       " ('baz', 'two'),\n",
       " ('foo', 'one'),\n",
       " ('foo', 'two'),\n",
       " ('qux', 'one'),\n",
       " ('qux', 'two')]"
      ]
     },
     "execution_count": 160,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tuples = list(zip(*[['bar', 'bar', 'baz', 'baz',\n",
    "                      'foo', 'foo', 'qux', 'qux'],\n",
    "                     ['one', 'two', 'one', 'two',\n",
    "                      'one', 'two', 'one', 'two']]))\n",
    "tuples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "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></th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">bar</th>\n",
       "      <th>one</th>\n",
       "      <td>0.779219</td>\n",
       "      <td>0.624401</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>two</th>\n",
       "      <td>0.643419</td>\n",
       "      <td>0.484125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">baz</th>\n",
       "      <th>one</th>\n",
       "      <td>0.293259</td>\n",
       "      <td>0.829503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>two</th>\n",
       "      <td>0.707516</td>\n",
       "      <td>0.181507</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">foo</th>\n",
       "      <th>one</th>\n",
       "      <td>0.819651</td>\n",
       "      <td>0.656431</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>two</th>\n",
       "      <td>0.958289</td>\n",
       "      <td>0.891413</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">qux</th>\n",
       "      <th>one</th>\n",
       "      <td>0.150244</td>\n",
       "      <td>0.217822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>two</th>\n",
       "      <td>0.313796</td>\n",
       "      <td>0.549416</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                A         B\n",
       "bar one  0.779219  0.624401\n",
       "    two  0.643419  0.484125\n",
       "baz one  0.293259  0.829503\n",
       "    two  0.707516  0.181507\n",
       "foo one  0.819651  0.656431\n",
       "    two  0.958289  0.891413\n",
       "qux one  0.150244  0.217822\n",
       "    two  0.313796  0.549416"
      ]
     },
     "execution_count": 165,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "index = pd.MultiIndex.from_tuples(tuples)\n",
    "df = pd.DataFrame(np.random.rand(8, 2), index=index, columns=['A', 'B'])\n",
    "df2 = df[:4]\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "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></th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">bar</th>\n",
       "      <th>one</th>\n",
       "      <td>0.779219</td>\n",
       "      <td>0.624401</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>two</th>\n",
       "      <td>0.643419</td>\n",
       "      <td>0.484125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">baz</th>\n",
       "      <th>one</th>\n",
       "      <td>0.293259</td>\n",
       "      <td>0.829503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>two</th>\n",
       "      <td>0.707516</td>\n",
       "      <td>0.181507</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                A         B\n",
       "bar one  0.779219  0.624401\n",
       "    two  0.643419  0.484125\n",
       "baz one  0.293259  0.829503\n",
       "    two  0.707516  0.181507"
      ]
     },
     "execution_count": 167,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bar  one  A    0.779219\n",
       "          B    0.624401\n",
       "     two  A    0.643419\n",
       "          B    0.484125\n",
       "baz  one  A    0.293259\n",
       "          B    0.829503\n",
       "     two  A    0.707516\n",
       "          B    0.181507\n",
       "dtype: float64"
      ]
     },
     "execution_count": 166,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stacked = df2.stack()\n",
    "stacked"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "pandas.core.series.Series"
      ]
     },
     "execution_count": 168,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(stacked)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "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></th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">bar</th>\n",
       "      <th>one</th>\n",
       "      <td>0.779219</td>\n",
       "      <td>0.624401</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>two</th>\n",
       "      <td>0.643419</td>\n",
       "      <td>0.484125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">baz</th>\n",
       "      <th>one</th>\n",
       "      <td>0.293259</td>\n",
       "      <td>0.829503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>two</th>\n",
       "      <td>0.707516</td>\n",
       "      <td>0.181507</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                A         B\n",
       "bar one  0.779219  0.624401\n",
       "    two  0.643419  0.484125\n",
       "baz one  0.293259  0.829503\n",
       "    two  0.707516  0.181507"
      ]
     },
     "execution_count": 169,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stacked.unstack()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "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></th>\n",
       "      <th>bar</th>\n",
       "      <th>baz</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">one</th>\n",
       "      <th>A</th>\n",
       "      <td>0.779219</td>\n",
       "      <td>0.293259</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>B</th>\n",
       "      <td>0.624401</td>\n",
       "      <td>0.829503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">two</th>\n",
       "      <th>A</th>\n",
       "      <td>0.643419</td>\n",
       "      <td>0.707516</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>B</th>\n",
       "      <td>0.484125</td>\n",
       "      <td>0.181507</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            bar       baz\n",
       "one A  0.779219  0.293259\n",
       "    B  0.624401  0.829503\n",
       "two A  0.643419  0.707516\n",
       "    B  0.484125  0.181507"
      ]
     },
     "execution_count": 171,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stacked.unstack(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "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></th>\n",
       "      <th>one</th>\n",
       "      <th>two</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">bar</th>\n",
       "      <th>A</th>\n",
       "      <td>0.779219</td>\n",
       "      <td>0.643419</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>B</th>\n",
       "      <td>0.624401</td>\n",
       "      <td>0.484125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">baz</th>\n",
       "      <th>A</th>\n",
       "      <td>0.293259</td>\n",
       "      <td>0.707516</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>B</th>\n",
       "      <td>0.829503</td>\n",
       "      <td>0.181507</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            one       two\n",
       "bar A  0.779219  0.643419\n",
       "    B  0.624401  0.484125\n",
       "baz A  0.293259  0.707516\n",
       "    B  0.829503  0.181507"
      ]
     },
     "execution_count": 172,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stacked.unstack(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pivot tables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "      <th>E</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>one</td>\n",
       "      <td>A</td>\n",
       "      <td>foo</td>\n",
       "      <td>-0.564908</td>\n",
       "      <td>1.041150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>one</td>\n",
       "      <td>B</td>\n",
       "      <td>foo</td>\n",
       "      <td>-1.402189</td>\n",
       "      <td>-0.194956</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>two</td>\n",
       "      <td>C</td>\n",
       "      <td>foo</td>\n",
       "      <td>0.214287</td>\n",
       "      <td>-1.223390</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>three</td>\n",
       "      <td>A</td>\n",
       "      <td>bar</td>\n",
       "      <td>1.196394</td>\n",
       "      <td>-0.637703</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>one</td>\n",
       "      <td>B</td>\n",
       "      <td>bar</td>\n",
       "      <td>-1.080359</td>\n",
       "      <td>-0.764453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>one</td>\n",
       "      <td>C</td>\n",
       "      <td>bar</td>\n",
       "      <td>-1.239668</td>\n",
       "      <td>0.483494</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>two</td>\n",
       "      <td>A</td>\n",
       "      <td>foo</td>\n",
       "      <td>0.074217</td>\n",
       "      <td>0.440609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>three</td>\n",
       "      <td>B</td>\n",
       "      <td>foo</td>\n",
       "      <td>0.958610</td>\n",
       "      <td>-0.044499</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>one</td>\n",
       "      <td>C</td>\n",
       "      <td>foo</td>\n",
       "      <td>0.857464</td>\n",
       "      <td>1.003873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>one</td>\n",
       "      <td>A</td>\n",
       "      <td>bar</td>\n",
       "      <td>-1.324573</td>\n",
       "      <td>-1.141071</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>two</td>\n",
       "      <td>B</td>\n",
       "      <td>bar</td>\n",
       "      <td>0.033825</td>\n",
       "      <td>-0.283092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>three</td>\n",
       "      <td>C</td>\n",
       "      <td>bar</td>\n",
       "      <td>-1.364382</td>\n",
       "      <td>-0.366823</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        A  B    C         D         E\n",
       "0     one  A  foo -0.564908  1.041150\n",
       "1     one  B  foo -1.402189 -0.194956\n",
       "2     two  C  foo  0.214287 -1.223390\n",
       "3   three  A  bar  1.196394 -0.637703\n",
       "4     one  B  bar -1.080359 -0.764453\n",
       "5     one  C  bar -1.239668  0.483494\n",
       "6     two  A  foo  0.074217  0.440609\n",
       "7   three  B  foo  0.958610 -0.044499\n",
       "8     one  C  foo  0.857464  1.003873\n",
       "9     one  A  bar -1.324573 -1.141071\n",
       "10    two  B  bar  0.033825 -0.283092\n",
       "11  three  C  bar -1.364382 -0.366823"
      ]
     },
     "execution_count": 173,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame({'A': ['one', 'one', 'two', 'three'] * 3,\n",
    "                    'B': ['A', 'B', 'C'] * 4,\n",
    "                    'C': ['foo', 'foo', 'foo', 'bar', 'bar', 'bar'] * 2,\n",
    "                    'D': np.random.randn(12),\n",
    "                    'E': np.random.randn(12)})\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "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>A</th>\n",
       "      <th>one</th>\n",
       "      <th>three</th>\n",
       "      <th>two</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">A</th>\n",
       "      <th>bar</th>\n",
       "      <td>-1.141071</td>\n",
       "      <td>-0.637703</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>foo</th>\n",
       "      <td>1.041150</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.440609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">B</th>\n",
       "      <th>bar</th>\n",
       "      <td>-0.764453</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.283092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>foo</th>\n",
       "      <td>-0.194956</td>\n",
       "      <td>-0.044499</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">C</th>\n",
       "      <th>bar</th>\n",
       "      <td>0.483494</td>\n",
       "      <td>-0.366823</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>foo</th>\n",
       "      <td>1.003873</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-1.223390</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "A           one     three       two\n",
       "B C                                \n",
       "A bar -1.141071 -0.637703       NaN\n",
       "  foo  1.041150       NaN  0.440609\n",
       "B bar -0.764453       NaN -0.283092\n",
       "  foo -0.194956 -0.044499       NaN\n",
       "C bar  0.483494 -0.366823       NaN\n",
       "  foo  1.003873       NaN -1.223390"
      ]
     },
     "execution_count": 175,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.pivot_table(df, index=['B', 'C'], columns='A', values='E')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## Time series"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DatetimeIndex(['2012-01-01 00:00:00', '2012-01-01 00:00:01',\n",
       "               '2012-01-01 00:00:02', '2012-01-01 00:00:03',\n",
       "               '2012-01-01 00:00:04'],\n",
       "              dtype='datetime64[ns]', freq='S')"
      ]
     },
     "execution_count": 178,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rng = pd.date_range('1/1/2012', periods=100, freq='S')\n",
    "rng[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2012-01-01 00:00:00    0.273328\n",
       "2012-01-01 00:00:01    0.194347\n",
       "2012-01-01 00:00:02    0.017551\n",
       "2012-01-01 00:00:03    0.266894\n",
       "2012-01-01 00:00:04    0.223220\n",
       "                         ...   \n",
       "2012-01-01 00:01:35    0.916800\n",
       "2012-01-01 00:01:36    0.571790\n",
       "2012-01-01 00:01:37    0.023535\n",
       "2012-01-01 00:01:38    0.090983\n",
       "2012-01-01 00:01:39    0.254118\n",
       "Freq: S, Length: 100, dtype: float64"
      ]
     },
     "execution_count": 182,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ts = pd.Series(np.random.rand(100), index=rng)\n",
    "ts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2012-01-01 00:00:00    29.688076\n",
       "2012-01-01 00:01:00    17.245419\n",
       "Freq: T, dtype: float64"
      ]
     },
     "execution_count": 184,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ts.resample('1min').sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## Categoricals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "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>id</th>\n",
       "      <th>raw_grade</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>b</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>b</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>e</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id raw_grade\n",
       "0   1         a\n",
       "1   2         b\n",
       "2   3         b\n",
       "3   4         a\n",
       "4   5         a\n",
       "5   6         e"
      ]
     },
     "execution_count": 185,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame({\"id\": [1, 2, 3, 4, 5, 6],\n",
    "                    \"raw_grade\": ['a', 'b', 'b', 'a', 'a', 'e']})\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    a\n",
       "1    b\n",
       "2    b\n",
       "3    a\n",
       "4    a\n",
       "5    e\n",
       "Name: grade, dtype: category\n",
       "Categories (3, object): ['a', 'b', 'e']"
      ]
     },
     "execution_count": 187,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['grade'] = df.raw_grade.astype(\"category\")\n",
    "df.grade"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       bad\n",
       "1    normal\n",
       "2    normal\n",
       "3       bad\n",
       "4       bad\n",
       "5      good\n",
       "Name: grade, dtype: category\n",
       "Categories (3, object): ['bad', 'normal', 'good']"
      ]
     },
     "execution_count": 188,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.grade.cat.categories = ['bad', 'normal', 'good']\n",
    "df.grade"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bad          3\n",
       "normal       2\n",
       "good         1\n",
       "very good    0\n",
       "Name: grade, dtype: int64"
      ]
     },
     "execution_count": 190,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['grade'] = df.grade.cat.set_categories(['bad', 'normal', 'good', 'very good'])\n",
    "df.grade.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "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>id</th>\n",
       "      <th>raw_grade</th>\n",
       "      <th>grade</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>a</td>\n",
       "      <td>bad</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>a</td>\n",
       "      <td>bad</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>a</td>\n",
       "      <td>bad</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>b</td>\n",
       "      <td>normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>b</td>\n",
       "      <td>normal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>e</td>\n",
       "      <td>good</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id raw_grade   grade\n",
       "0   1         a     bad\n",
       "3   4         a     bad\n",
       "4   5         a     bad\n",
       "1   2         b  normal\n",
       "2   3         b  normal\n",
       "5   6         e    good"
      ]
     },
     "execution_count": 193,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sort_values(by='grade')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plotting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2000-01-01    0.383413\n",
       "2000-01-02   -0.141648\n",
       "2000-01-03   -0.329732\n",
       "2000-01-04    0.433057\n",
       "2000-01-05    0.100554\n",
       "Freq: D, dtype: float64"
      ]
     },
     "execution_count": 202,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ts = pd.Series(np.random.randn(1000),\n",
    "                index=pd.date_range('1/1/2000', periods=1000))\n",
    "ts[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 203,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "ts.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 204,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "ts.cumsum().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2000-01-01</th>\n",
       "      <td>1.261520</td>\n",
       "      <td>0.167444</td>\n",
       "      <td>0.777502</td>\n",
       "      <td>-0.908136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-01-02</th>\n",
       "      <td>0.484081</td>\n",
       "      <td>1.122647</td>\n",
       "      <td>-0.489198</td>\n",
       "      <td>1.310879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-01-03</th>\n",
       "      <td>0.764666</td>\n",
       "      <td>1.319598</td>\n",
       "      <td>1.176613</td>\n",
       "      <td>-0.944798</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-01-04</th>\n",
       "      <td>1.378487</td>\n",
       "      <td>1.888987</td>\n",
       "      <td>-0.680342</td>\n",
       "      <td>-0.752147</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-01-05</th>\n",
       "      <td>0.894348</td>\n",
       "      <td>-1.061797</td>\n",
       "      <td>2.110745</td>\n",
       "      <td>1.074145</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2002-09-22</th>\n",
       "      <td>0.810336</td>\n",
       "      <td>-0.803027</td>\n",
       "      <td>-1.087350</td>\n",
       "      <td>-0.757818</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2002-09-23</th>\n",
       "      <td>1.702490</td>\n",
       "      <td>1.862966</td>\n",
       "      <td>-1.246053</td>\n",
       "      <td>-0.500768</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2002-09-24</th>\n",
       "      <td>-1.115860</td>\n",
       "      <td>0.157061</td>\n",
       "      <td>0.416421</td>\n",
       "      <td>-1.579590</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2002-09-25</th>\n",
       "      <td>0.128416</td>\n",
       "      <td>-0.586443</td>\n",
       "      <td>-0.736478</td>\n",
       "      <td>-0.498537</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2002-09-26</th>\n",
       "      <td>0.300037</td>\n",
       "      <td>-2.647326</td>\n",
       "      <td>1.029755</td>\n",
       "      <td>0.390894</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "2000-01-01  1.261520  0.167444  0.777502 -0.908136\n",
       "2000-01-02  0.484081  1.122647 -0.489198  1.310879\n",
       "2000-01-03  0.764666  1.319598  1.176613 -0.944798\n",
       "2000-01-04  1.378487  1.888987 -0.680342 -0.752147\n",
       "2000-01-05  0.894348 -1.061797  2.110745  1.074145\n",
       "...              ...       ...       ...       ...\n",
       "2002-09-22  0.810336 -0.803027 -1.087350 -0.757818\n",
       "2002-09-23  1.702490  1.862966 -1.246053 -0.500768\n",
       "2002-09-24 -1.115860  0.157061  0.416421 -1.579590\n",
       "2002-09-25  0.128416 -0.586443 -0.736478 -0.498537\n",
       "2002-09-26  0.300037 -2.647326  1.029755  0.390894\n",
       "\n",
       "[1000 rows x 4 columns]"
      ]
     },
     "execution_count": 205,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame(np.random.randn(1000, 4), index=ts.index,\n",
    "                   columns=['A', 'B', 'C', 'D'])\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 206,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "df.cumsum().plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## Getting data in/out"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## Gotchas"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "# Gotchas / FAQ"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- https://pandas.pydata.org/docs/user_guide/gotchas.html#gotchas"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Cookbook"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>10</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>20</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>30</td>\n",
       "      <td>-30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7</td>\n",
       "      <td>40</td>\n",
       "      <td>-50</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC\n",
       "0    4   10  100\n",
       "1    5   20   50\n",
       "2    6   30  -30\n",
       "3    7   40  -50"
      ]
     },
     "execution_count": 209,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame({'AAA': [4, 5, 6, 7],\n",
    "                       'BBB': [10, 20, 30, 40],\n",
    "                       'CCC': [100, 50, -30, -50]})\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## Idioms"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "### if-then columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>999</td>\n",
       "      <td>-30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7</td>\n",
       "      <td>999</td>\n",
       "      <td>-50</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC\n",
       "0    4    0  100\n",
       "1    5    0   50\n",
       "2    6  999  -30\n",
       "3    7  999  -50"
      ]
     },
     "execution_count": 210,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.loc[df.AAA <= 5, ['BBB']] = 0\n",
    "df.loc[df.AAA > 5, ['BBB']] = 999\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>999</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>999</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>-30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>-50</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC\n",
       "0    4  999  100\n",
       "1    5  999   50\n",
       "2    6    0  -30\n",
       "3    7    0  -50"
      ]
     },
     "execution_count": 212,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['BBB'] = df.apply(lambda x: 999 if x.AAA <= 5 else 0, axis=1)\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>999</td>\n",
       "      <td>-30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7</td>\n",
       "      <td>999</td>\n",
       "      <td>-50</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC\n",
       "0    4    0  100\n",
       "1    5    0   50\n",
       "2    6  999  -30\n",
       "3    7  999  -50"
      ]
     },
     "execution_count": 221,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['BBB'] = np.where(df.BBB == 999, 0, 999)\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1000</td>\n",
       "      <td>999</td>\n",
       "      <td>-1000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1000</td>\n",
       "      <td>999</td>\n",
       "      <td>-1000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-1000</td>\n",
       "      <td>-1000</td>\n",
       "      <td>-1000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-1000</td>\n",
       "      <td>-1000</td>\n",
       "      <td>-1000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    AAA   BBB   CCC\n",
       "0 -1000   999 -1000\n",
       "1 -1000   999 -1000\n",
       "2 -1000 -1000 -1000\n",
       "3 -1000 -1000 -1000"
      ]
     },
     "execution_count": 218,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.where(df >= 999, -1000)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Splitting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>999</td>\n",
       "      <td>-30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7</td>\n",
       "      <td>999</td>\n",
       "      <td>-50</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC\n",
       "0    4    0  100\n",
       "1    5    0   50\n",
       "2    6  999  -30\n",
       "3    7  999  -50"
      ]
     },
     "execution_count": 222,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC\n",
       "0    4    0  100\n",
       "1    5    0   50"
      ]
     },
     "execution_count": 223,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1 = df[df.AAA <= 5]\n",
    "df1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "### Building criteria"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>10</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>20</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>30</td>\n",
       "      <td>-30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7</td>\n",
       "      <td>40</td>\n",
       "      <td>-50</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC\n",
       "0    4   10  100\n",
       "1    5   20   50\n",
       "2    6   30  -30\n",
       "3    7   40  -50"
      ]
     },
     "execution_count": 240,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame({'AAA': [4, 5, 6, 7],\n",
    "                      'BBB': [10, 20, 30, 40],\n",
    "                      'CCC': [100, 50, -30, -50]})\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 選最接近某指定值的 row"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "17.5"
      ]
     },
     "execution_count": 241,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "c_mean = df['CCC'].mean()\n",
    "c_mean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "      <th>abs</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>20</td>\n",
       "      <td>50</td>\n",
       "      <td>32.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>30</td>\n",
       "      <td>-30</td>\n",
       "      <td>47.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7</td>\n",
       "      <td>40</td>\n",
       "      <td>-50</td>\n",
       "      <td>67.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>10</td>\n",
       "      <td>100</td>\n",
       "      <td>82.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC   abs\n",
       "1    5   20   50  32.5\n",
       "2    6   30  -30  47.5\n",
       "3    7   40  -50  67.5\n",
       "0    4   10  100  82.5"
      ]
     },
     "execution_count": 243,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['abs'] = (df.CCC - c_mean).abs().values\n",
    "df.sort_values('abs')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "      <th>abs</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>20</td>\n",
       "      <td>50</td>\n",
       "      <td>32.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>30</td>\n",
       "      <td>-30</td>\n",
       "      <td>47.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7</td>\n",
       "      <td>40</td>\n",
       "      <td>-50</td>\n",
       "      <td>67.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>10</td>\n",
       "      <td>100</td>\n",
       "      <td>82.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC   abs\n",
       "1    5   20   50  32.5\n",
       "2    6   30  -30  47.5\n",
       "3    7   40  -50  67.5\n",
       "0    4   10  100  82.5"
      ]
     },
     "execution_count": 245,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.loc[(df.CCC - c_mean).abs().argsort()]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 從多個 filters 動態建立最終 filter "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {},
   "outputs": [],
   "source": [
    "import functools"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "      <th>abs</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>20</td>\n",
       "      <td>50</td>\n",
       "      <td>32.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC   abs\n",
       "1    5   20   50  32.5"
      ]
     },
     "execution_count": 248,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f1 = df.AAA >= 5\n",
    "f2 = df.BBB <= 20\n",
    "f3 = df.CCC >= 50\n",
    "df[f1 & f2 & f3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "      <th>abs</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>20</td>\n",
       "      <td>50</td>\n",
       "      <td>32.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC   abs\n",
       "1    5   20   50  32.5"
      ]
     },
     "execution_count": 250,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f = functools.reduce(lambda x, y: x & y, [f1, f2, f3])\n",
    "df[f]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "      <th>abs</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>10</td>\n",
       "      <td>100</td>\n",
       "      <td>82.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>20</td>\n",
       "      <td>50</td>\n",
       "      <td>32.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>30</td>\n",
       "      <td>-30</td>\n",
       "      <td>47.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7</td>\n",
       "      <td>40</td>\n",
       "      <td>-50</td>\n",
       "      <td>67.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC   abs\n",
       "0    4   10  100  82.5\n",
       "1    5   20   50  32.5\n",
       "2    6   30  -30  47.5\n",
       "3    7   40  -50  67.5"
      ]
     },
     "execution_count": 251,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 同時 label + positional 選"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "      <th>abs</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6</td>\n",
       "      <td>30</td>\n",
       "      <td>-30</td>\n",
       "      <td>47.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7</td>\n",
       "      <td>40</td>\n",
       "      <td>-50</td>\n",
       "      <td>67.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC   abs\n",
       "2    6   30  -30  47.5\n",
       "3    7   40  -50  67.5"
      ]
     },
     "execution_count": 252,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[(df.AAA >= 5) & (df.index.isin([2, 3]))]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Ambiguity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4</td>\n",
       "      <td>10</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5</td>\n",
       "      <td>20</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>6</td>\n",
       "      <td>30</td>\n",
       "      <td>-30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>7</td>\n",
       "      <td>40</td>\n",
       "      <td>-50</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC\n",
       "1    4   10  100\n",
       "2    5   20   50\n",
       "3    6   30  -30\n",
       "4    7   40  -50"
      ]
     },
     "execution_count": 253,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = {'AAA': [4, 5, 6, 7],\n",
    "            'BBB': [10, 20, 30, 40],\n",
    "           'CCC': [100, 50, -30, -50]}\n",
    "df2 = pd.DataFrame(data=data, index=[1, 2, 3, 4])\n",
    "df2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5</td>\n",
       "      <td>20</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>6</td>\n",
       "      <td>30</td>\n",
       "      <td>-30</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC\n",
       "2    5   20   50\n",
       "3    6   30  -30"
      ]
     },
     "execution_count": 254,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.iloc[1:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4</td>\n",
       "      <td>10</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5</td>\n",
       "      <td>20</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>6</td>\n",
       "      <td>30</td>\n",
       "      <td>-30</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC\n",
       "1    4   10  100\n",
       "2    5   20   50\n",
       "3    6   30  -30"
      ]
     },
     "execution_count": 255,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.loc[1:3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 用 apply /applymap 動態加入新欄位"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC\n",
       "0    1    1    2\n",
       "1    2    1    1\n",
       "2    1    2    3\n",
       "3    3    2    1"
      ]
     },
     "execution_count": 257,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame({'AAA': [1, 2, 1, 3],\n",
    "                       'BBB': [1, 1, 2, 2],\n",
    "                       'CCC': [2, 1, 3, 1]})\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "      <th>AAA_cat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>apple</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>banana</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>apple</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>cat</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC AAA_cat\n",
       "0    1    1    2   apple\n",
       "1    2    1    1  banana\n",
       "2    1    2    3   apple\n",
       "3    3    2    1     cat"
      ]
     },
     "execution_count": 267,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cates = dict(zip(range(1, 4), ['apple', 'banana', 'cat']))\n",
    "df['AAA_cat'] = df[['AAA']].applymap(cates.get)\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 269,
   "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>AAA</th>\n",
       "      <th>BBB</th>\n",
       "      <th>CCC</th>\n",
       "      <th>AAA_cat</th>\n",
       "      <th>AAA_cat2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>apple</td>\n",
       "      <td>apple</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>banana</td>\n",
       "      <td>banana</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>apple</td>\n",
       "      <td>apple</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>cat</td>\n",
       "      <td>cat</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA  BBB  CCC AAA_cat AAA_cat2\n",
       "0    1    1    2   apple    apple\n",
       "1    2    1    1  banana   banana\n",
       "2    1    2    3   apple    apple\n",
       "3    3    2    1     cat      cat"
      ]
     },
     "execution_count": 269,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['AAA_cat2'] = df.AAA.apply(lambda x: cates.get(x))\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 261,
   "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>AAA</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA\n",
       "0    1\n",
       "1    4\n",
       "2    1\n",
       "3    9"
      ]
     },
     "execution_count": 261,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[['AAA']].applymap(lambda x: x**2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 262,
   "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>AAA</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   AAA\n",
       "0    1\n",
       "1    4\n",
       "2    1\n",
       "3    9"
      ]
     },
     "execution_count": 262,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[['AAA']].apply(lambda x: x**2, axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DataCamp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "# Blog"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## 建立 DataFrame"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "pandas 裡有非常多種可以初始化一個 [DataFrame](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.html) 的技巧。以下列出一些我覺得實用的初始化方式。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 用 Python dict 建立 DataFrame"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "使用 Python 的 `dict` 來初始化 DataFrame 十分直覺。基本上 `dict` 裡頭的每一個鍵值（key）都對應到一個欄位名稱，而其值（value）則是一個 iterable，代表該欄位裡頭所有的數值。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>col 1</th>\n",
       "      <th>col 2</th>\n",
       "      <th>col 3</th>\n",
       "      <th>col 4</th>\n",
       "      <th>col 5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "      <td>x</td>\n",
       "      <td>a</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>20</td>\n",
       "      <td>y</td>\n",
       "      <td>b</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>30</td>\n",
       "      <td>z</td>\n",
       "      <td>c</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   col 1  col 2 col 3 col 4  col 5\n",
       "0      1     10     x     a      0\n",
       "1      2     20     y     b      1\n",
       "2      3     30     z     c      2"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dic = {\n",
    "    \"col 1\": [1, 2, 3], \n",
    "    \"col 2\": [10, 20, 30],\n",
    "    \"col 3\": list('xyz'),\n",
    "    \"col 4\": ['a', 'b', 'c'],\n",
    "    \"col 5\": pd.Series(range(3))\n",
    "}\n",
    "df = pd.DataFrame(dic)\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在需要管理多個 DataFrames 時你會想要用更有意義的名字來代表它們，但在資料科學領域裡只要看到 `df`，每個人都會預期它是一個 **D**ata**F**rame，不論是 Python 或是 R 語言的使用者。\n",
    "\n",
    "很多時候你也會需要改變 DataFrame 裡的欄位名稱："
   ]
  },
  {
   "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>x</th>\n",
       "      <th>10x</th>\n",
       "      <th>col 3</th>\n",
       "      <th>col 4</th>\n",
       "      <th>col 5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "      <td>x</td>\n",
       "      <td>a</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>20</td>\n",
       "      <td>y</td>\n",
       "      <td>b</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>30</td>\n",
       "      <td>z</td>\n",
       "      <td>c</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   x  10x col 3 col 4  col 5\n",
       "0  1   10     x     a      0\n",
       "1  2   20     y     b      1\n",
       "2  3   30     z     c      2"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rename_dic = {\"col 1\": \"x\", \"col 2\": \"10x\"}\n",
    "df.rename(rename_dic, axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "這邊也很直覺，就是給一個將舊欄位名對應到新欄位名的 Python `dict`。值得注意的是參數 `axis=1`：在 pandas 裡大部分函式預設處理的軸為列（row）：以 `axis=0` 表示；而將 `axis` 設置為 `1` 則代表你想以行（column）為單位套用該函式。\n",
    "\n",
    "你也可以用 `df.columns` 的方式改欄位名稱："
   ]
  },
  {
   "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>x(new)</th>\n",
       "      <th>10x(new)</th>\n",
       "      <th>col 3</th>\n",
       "      <th>col 4</th>\n",
       "      <th>col 5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "      <td>x</td>\n",
       "      <td>a</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>20</td>\n",
       "      <td>y</td>\n",
       "      <td>b</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>30</td>\n",
       "      <td>z</td>\n",
       "      <td>c</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   x(new)  10x(new) col 3 col 4  col 5\n",
       "0       1        10     x     a      0\n",
       "1       2        20     y     b      1\n",
       "2       3        30     z     c      2"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns = ['x(new)', '10x(new)'] + list(df.columns[2:])\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 使用 pd.util.testing 隨機建立 DataFrame"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "當你想要隨意初始化一個 DataFrame 並測試 pandas 功能時，`pd.util.testing` 就顯得十分好用："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>xPlIXbQqGU</th>\n",
       "      <td>1.086581</td>\n",
       "      <td>-0.002484</td>\n",
       "      <td>-0.335693</td>\n",
       "      <td>0.226988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>IAFe6K8mpA</th>\n",
       "      <td>-0.547556</td>\n",
       "      <td>-0.290935</td>\n",
       "      <td>-0.014313</td>\n",
       "      <td>-0.301007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OoATGY0k2M</th>\n",
       "      <td>1.017636</td>\n",
       "      <td>0.568835</td>\n",
       "      <td>-0.272382</td>\n",
       "      <td>0.659657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>uRN2yGacDw</th>\n",
       "      <td>-0.662390</td>\n",
       "      <td>1.929820</td>\n",
       "      <td>-1.206670</td>\n",
       "      <td>0.250626</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ElphZli9nK</th>\n",
       "      <td>-0.697235</td>\n",
       "      <td>0.942415</td>\n",
       "      <td>-0.894887</td>\n",
       "      <td>0.701790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>oiEoCPCXK8</th>\n",
       "      <td>-1.049284</td>\n",
       "      <td>-1.019107</td>\n",
       "      <td>-0.640271</td>\n",
       "      <td>-0.613056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NUrQFrYQw1</th>\n",
       "      <td>0.759355</td>\n",
       "      <td>0.717367</td>\n",
       "      <td>-0.449368</td>\n",
       "      <td>1.889321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>oC9iiEBneW</th>\n",
       "      <td>0.665412</td>\n",
       "      <td>-0.391204</td>\n",
       "      <td>-0.974010</td>\n",
       "      <td>0.248326</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4hD6Eea7yF</th>\n",
       "      <td>-0.862819</td>\n",
       "      <td>2.092149</td>\n",
       "      <td>0.976645</td>\n",
       "      <td>-0.388735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3QD5mMfstw</th>\n",
       "      <td>-0.312762</td>\n",
       "      <td>-0.110278</td>\n",
       "      <td>1.162421</td>\n",
       "      <td>-0.335144</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "xPlIXbQqGU  1.086581 -0.002484 -0.335693  0.226988\n",
       "IAFe6K8mpA -0.547556 -0.290935 -0.014313 -0.301007\n",
       "OoATGY0k2M  1.017636  0.568835 -0.272382  0.659657\n",
       "uRN2yGacDw -0.662390  1.929820 -1.206670  0.250626\n",
       "ElphZli9nK -0.697235  0.942415 -0.894887  0.701790\n",
       "oiEoCPCXK8 -1.049284 -1.019107 -0.640271 -0.613056\n",
       "NUrQFrYQw1  0.759355  0.717367 -0.449368  1.889321\n",
       "oC9iiEBneW  0.665412 -0.391204 -0.974010  0.248326\n",
       "4hD6Eea7yF -0.862819  2.092149  0.976645 -0.388735\n",
       "3QD5mMfstw -0.312762 -0.110278  1.162421 -0.335144"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.util.testing.makeDataFrame().head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`head` 函式預設用來顯示 DataFrame 中前 5 筆數據。要顯示後面數據則可以使用 `tail` 函式。\n",
    "\n",
    "你也可以用 `makeMixedDataFrame` 建立一個有各種資料型態的 DataFrame 方便測試："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>foo1</td>\n",
       "      <td>2009-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>foo2</td>\n",
       "      <td>2009-01-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>foo3</td>\n",
       "      <td>2009-01-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>foo4</td>\n",
       "      <td>2009-01-06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>foo5</td>\n",
       "      <td>2009-01-07</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     A    B     C          D\n",
       "0  0.0  0.0  foo1 2009-01-01\n",
       "1  1.0  1.0  foo2 2009-01-02\n",
       "2  2.0  0.0  foo3 2009-01-05\n",
       "3  3.0  1.0  foo4 2009-01-06\n",
       "4  4.0  0.0  foo5 2009-01-07"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.util.testing.makeMixedDataFrame()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "其他函式如 `makeMissingDataframe` 及 `makeTimeDataFrame` 在後面的章節都還會看到。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 將剪貼簿內容轉換成 DataFrame"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你可以從 Excel、Google Sheet 或是網頁上複製表格並將其轉成 DataFrame。\n",
    "\n",
    "簡單 2 步驟：\n",
    "- 複製其他來源的表格\n",
    "- 執行 `pd.read_clipboard`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!mp4\n",
    "- dark\n",
    "- images/pandas/pandas_clipboard.mp4\n",
    "- images/pandas/pandas_clipboard.jpg"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "這個技巧在你想要快速將一些數據轉成 DataFrame 時非常方便。當然，你得考量重現性（reproducibility）。\n",
    "\n",
    "為了讓未來的自己以及他人可以重現你當下的結果，必要時記得另存新檔以供後人使用：\n",
    "\n",
    "```python\n",
    "df.to_csv(\"some_data.csv\")\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 讀取線上 CSV 檔"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "不限於本地檔案，只要有正確的 URL 以及網路連線就可以將網路上的任意 CSV 檔案轉成 DataFrame。\n",
    "\n",
    "比方說你可以將 Kaggle 著名的[鐵達尼號競賽](https://www.kaggle.com/c/titanic)的 CSV 檔案從網路上下載下來並轉成 DataFrame："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Braund, Mr. Owen Harris</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Cumings, Mrs. John Bradley (Florence Briggs Th...</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Heikkinen, Miss. Laina</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Futrelle, Mrs. Jacques Heath (Lily May Peel)</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Allen, Mr. William Henry</td>\n",
       "      <td>male</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>373450</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived  Pclass  \\\n",
       "0            1         0       3   \n",
       "1            2         1       1   \n",
       "2            3         1       3   \n",
       "3            4         1       1   \n",
       "4            5         0       3   \n",
       "\n",
       "                                                Name     Sex   Age  SibSp  \\\n",
       "0                            Braund, Mr. Owen Harris    male  22.0      1   \n",
       "1  Cumings, Mrs. John Bradley (Florence Briggs Th...  female  38.0      1   \n",
       "2                             Heikkinen, Miss. Laina  female  26.0      0   \n",
       "3       Futrelle, Mrs. Jacques Heath (Lily May Peel)  female  35.0      1   \n",
       "4                           Allen, Mr. William Henry    male  35.0      0   \n",
       "\n",
       "   Parch            Ticket     Fare Cabin Embarked  \n",
       "0      0         A/5 21171   7.2500   NaN        S  \n",
       "1      0          PC 17599  71.2833   C85        C  \n",
       "2      0  STON/O2. 3101282   7.9250   NaN        S  \n",
       "3      0            113803  53.1000  C123        S  \n",
       "4      0            373450   8.0500   NaN        S  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('http://bit.ly/kaggletrain')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "以下則是另個使用 pandas 爬取網路上數據並作分析的真實案例。\n",
    "\n",
    "我在之前的[ Chartify 教學文](https://leemeng.tw/chartify-a-simple-yet-powerful-python-data-visualization-tool-which-boost-your-productivity-as-a-data-scientist.html)中從臺北市的資料開放平台爬取[ A1 及 A2 類交通事故數據](https://data.taipei/#/dataset/detail?id=2f238b4f-1b27-4085-93e9-d684ef0e2735)後做了簡單匯總："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 271,
   "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>發生年</th>\n",
       "      <th>發生月</th>\n",
       "      <th>發生日</th>\n",
       "      <th>發生時</th>\n",
       "      <th>發生分</th>\n",
       "      <th>處理別</th>\n",
       "      <th>區序</th>\n",
       "      <th>肇事地點</th>\n",
       "      <th>死亡人數</th>\n",
       "      <th>受傷人數</th>\n",
       "      <th>當事人序</th>\n",
       "      <th>車種</th>\n",
       "      <th>性別</th>\n",
       "      <th>年齡</th>\n",
       "      <th>受傷程度</th>\n",
       "      <th>4天候</th>\n",
       "      <th>7速限</th>\n",
       "      <th>8道路型態</th>\n",
       "      <th>9事故位置</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>107</td>\n",
       "      <td>3</td>\n",
       "      <td>29</td>\n",
       "      <td>15</td>\n",
       "      <td>54</td>\n",
       "      <td>2</td>\n",
       "      <td>01大同區</td>\n",
       "      <td>大同區民權西路108號</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>B03</td>\n",
       "      <td>2</td>\n",
       "      <td>41.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>8</td>\n",
       "      <td>50</td>\n",
       "      <td>14.0</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>107</td>\n",
       "      <td>3</td>\n",
       "      <td>29</td>\n",
       "      <td>15</td>\n",
       "      <td>54</td>\n",
       "      <td>2</td>\n",
       "      <td>01大同區</td>\n",
       "      <td>大同區民權西路108號</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>C03</td>\n",
       "      <td>2</td>\n",
       "      <td>58.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8</td>\n",
       "      <td>50</td>\n",
       "      <td>14.0</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>107</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>17</td>\n",
       "      <td>42</td>\n",
       "      <td>2</td>\n",
       "      <td>01大同區</td>\n",
       "      <td>大同區重慶北路2段與南京西路口</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>B01</td>\n",
       "      <td>1</td>\n",
       "      <td>59.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>6</td>\n",
       "      <td>40</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>107</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>17</td>\n",
       "      <td>42</td>\n",
       "      <td>2</td>\n",
       "      <td>01大同區</td>\n",
       "      <td>大同區重慶北路3段與南京西路口</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>C03</td>\n",
       "      <td>1</td>\n",
       "      <td>18.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>6</td>\n",
       "      <td>40</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>107</td>\n",
       "      <td>1</td>\n",
       "      <td>14</td>\n",
       "      <td>9</td>\n",
       "      <td>56</td>\n",
       "      <td>2</td>\n",
       "      <td>01大同區</td>\n",
       "      <td>大同區承德路3段與民族西路口</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>C03</td>\n",
       "      <td>1</td>\n",
       "      <td>20.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8</td>\n",
       "      <td>50</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   發生年  發生月  發生日  發生時  發生分  處理別     區序             肇事地點  死亡人數  受傷人數  當事人序  \\\n",
       "0  107    3   29   15   54    2  01大同區      大同區民權西路108號     0     1     1   \n",
       "1  107    3   29   15   54    2  01大同區      大同區民權西路108號     0     1     2   \n",
       "2  107    1    7   17   42    2  01大同區  大同區重慶北路2段與南京西路口     0     1     1   \n",
       "3  107    1    7   17   42    2  01大同區  大同區重慶北路3段與南京西路口     0     1     2   \n",
       "4  107    1   14    9   56    2  01大同區   大同區承德路3段與民族西路口     0     1     1   \n",
       "\n",
       "    車種  性別    年齡  受傷程度  4天候  7速限  8道路型態  9事故位置  \n",
       "0  B03   2  41.0   3.0    8   50   14.0    9.0  \n",
       "1  C03   2  58.0   2.0    8   50   14.0    9.0  \n",
       "2  B01   1  59.0   3.0    6   40    4.0    2.0  \n",
       "3  C03   1  18.0   2.0    6   40    4.0    2.0  \n",
       "4  C03   1  20.0   2.0    8   50    4.0    1.0  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>區序</th>\n",
       "      <th>性別</th>\n",
       "      <th>死亡人數</th>\n",
       "      <th>受傷人數</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>南港區</td>\n",
       "      <td>女性</td>\n",
       "      <td>2</td>\n",
       "      <td>799</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>萬華區</td>\n",
       "      <td>女性</td>\n",
       "      <td>2</td>\n",
       "      <td>1084</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>松山區</td>\n",
       "      <td>女性</td>\n",
       "      <td>30</td>\n",
       "      <td>1146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>大同區</td>\n",
       "      <td>女性</td>\n",
       "      <td>3</td>\n",
       "      <td>1336</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>士林區</td>\n",
       "      <td>女性</td>\n",
       "      <td>3</td>\n",
       "      <td>1660</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    區序  性別  死亡人數  受傷人數\n",
       "0  南港區  女性     2   799\n",
       "1  萬華區  女性     2  1084\n",
       "2  松山區  女性    30  1146\n",
       "3  大同區  女性     3  1336\n",
       "4  士林區  女性     3  1660"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 你可以用類似的方式爬取任何網路上的公開數據集\n",
    "base_url = \"https://data.taipei/api/getDatasetInfo/downloadResource?id={}&rid={}\"\n",
    "_id = \"2f238b4f-1b27-4085-93e9-d684ef0e2735\"\n",
    "rid = \"ea731a84-e4a1-4523-b981-b733beddbc1f\"\n",
    "csv_url = base_url.format(_id, rid)\n",
    "df_raw = pd.read_csv(csv_url, encoding='big5')\n",
    "\n",
    "# 複製一份做處理\n",
    "df = df_raw.copy()\n",
    "\n",
    "# 計算不同區不同性別的死亡、受傷人數\n",
    "df['區序'] = df['區序'].apply(lambda x: ''.join([i for i in x if not i.isdigit()]))\n",
    "df = (df[df['性別'].isin([1, 2])]\n",
    "      .groupby(['區序', '性別'])[['死亡人數', '受傷人數']]\n",
    "      .sum()\n",
    "      .reset_index()\n",
    "      .sort_values('受傷人數'))\n",
    "\n",
    "df['性別'] = df['性別'].apply(lambda x: '男性' if x == 1 else '女性')\n",
    "df = df.reset_index().drop('index', axis=1)\n",
    "\n",
    "# 顯示結果\n",
    "display(df_raw.head())\n",
    "display(df.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "有了匯總過後的 DataFrame，你可以用後面[簡易繪圖並修改預設樣式](#簡易繪圖並修改預設樣式)章節提到的 pandas `plot` 函式繪圖。但在這邊讓我用 [Chartify](https://leemeng.tw/chartify-a-simple-yet-powerful-python-data-visualization-tool-which-boost-your-productivity-as-a-data-scientist.html) 展示結果：\n",
    "\n",
    "```python\n",
    "import chartify\n",
    "ch = chartify.Chart(\n",
    "    x_axis_type='categorical', \n",
    "    y_axis_type='categorical')\n",
    "\n",
    "ch.plot.heatmap(\n",
    "    data_frame=df, \n",
    "    y_column='性別', \n",
    "    x_column='區序', \n",
    "    color_column='受傷人數', \n",
    "    text_column='受傷人數', \n",
    "    color_palette='Reds', \n",
    "    text_format='{:,.0f}')\n",
    "\n",
    "(ch.set_title('2017 年台北市各行政區交通意外受傷人數')\n",
    " .set_subtitle('性別分計')\n",
    " .set_source_label(\"Data.Taipei\")\n",
    " .axes.set_xaxis_label('性別')\n",
    " .axes.set_yaxis_label('行政區')\n",
    " .show('png'))\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!image\n",
    "- pandas/read_csv_and_render_using_chartify.png"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "過來人經驗。雖然像這樣利用 pandas 直接從網路上下載並分析數據很方便，有時 host 數據的網頁與機構（尤其是政府機關）會無預期地修改他們網站，導致數據集的 URL 失效（苦主）。為了最大化重現性，我還是會建議將數據載到本地備份之後，再做分析比較實在。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 優化記憶體使用量"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你可以透過 `df.info` 查看 DataFrame 當前的記憶體用量："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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: 322.0 KB\n"
     ]
    }
   ],
   "source": [
    "df.info(memory_usage=\"deep\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "從最後一列可以看出鐵達尼號這個小 DataFrame 只佔了 322 KB。如果你是透過 [Jupyter 筆記本](https://jupyter.org/)來操作 pandas，也可以考慮用 [Variable Inspector](https://jupyter-contrib-nbextensions.readthedocs.io/en/latest/nbextensions/varInspector/README.html) 插件來觀察包含 DataFrame 等變數的大小："
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!mp4\n",
    "- dark\n",
    "- images/pandas/variable_inspector.mp4\n",
    "- Variable Inspector"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "這邊使用的 `df` 不佔什麼記憶體，但如果你想讀入的 DataFrame 很大，可以只讀入特定的欄位並將已知的分類型（categorical）欄位轉成 `category` 型態以節省記憶體（在分類數目較數據量小時有效）："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 891 entries, 0 to 890\n",
      "Data columns (total 4 columns):\n",
      "PassengerId    891 non-null int64\n",
      "Name           891 non-null object\n",
      "Sex            891 non-null object\n",
      "Embarked       889 non-null category\n",
      "dtypes: category(1), int64(1), object(2)\n",
      "memory usage: 134.9 KB\n"
     ]
    }
   ],
   "source": [
    "dtypes = {\"Embarked\": \"category\"}\n",
    "cols = ['PassengerId', 'Name', 'Sex', 'Embarked']\n",
    "df = pd.read_csv('http://bit.ly/kaggletrain', \n",
    "                 dtype=dtypes, usecols=cols)\n",
    "df.info(memory_usage=\"deep\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "透過減少讀入的欄位數並將 `object` 轉換成 `category` 欄位，讀入的 `df` 只剩 135 KB。只需剛剛的 40 % 記憶體用量。\n",
    "\n",
    "另外如果你想在有限的記憶體內處理巨大 CSV 檔案，也可以透過 `chunksize` 參數來限制一次讀入的列數（rows）："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>Braund, Mr. Owen Harris</td>\n",
       "      <td>male</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>Cumings, Mrs. John Bradley (Florence Briggs Th...</td>\n",
       "      <td>female</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>Heikkinen, Miss. Laina</td>\n",
       "      <td>female</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>Futrelle, Mrs. Jacques Heath (Lily May Peel)</td>\n",
       "      <td>female</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId                                               Name     Sex  \\\n",
       "0            1                            Braund, Mr. Owen Harris    male   \n",
       "1            2  Cumings, Mrs. John Bradley (Florence Briggs Th...  female   \n",
       "2            3                             Heikkinen, Miss. Laina  female   \n",
       "3            4       Futrelle, Mrs. Jacques Heath (Lily May Peel)  female   \n",
       "\n",
       "  Embarked  \n",
       "0        S  \n",
       "1        C  \n",
       "2        S  \n",
       "3        S  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>Allen, Mr. William Henry</td>\n",
       "      <td>male</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>Moran, Mr. James</td>\n",
       "      <td>male</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>McCarthy, Mr. Timothy J</td>\n",
       "      <td>male</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8</td>\n",
       "      <td>Palsson, Master. Gosta Leonard</td>\n",
       "      <td>male</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId                            Name   Sex Embarked\n",
       "4            5        Allen, Mr. William Henry  male        S\n",
       "5            6                Moran, Mr. James  male        Q\n",
       "6            7         McCarthy, Mr. Timothy J  male        S\n",
       "7            8  Palsson, Master. Gosta Leonard  male        S"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from IPython.display import display\n",
    "\n",
    "# chunksize=4 表示一次讀入 4 筆樣本\n",
    "reader = pd.read_csv('dataset/titanic-train.csv', \n",
    "                     chunksize=4, usecols=cols)\n",
    "# 秀出前兩個 chunks\n",
    "for _, df_partial in zip(range(2), reader):\n",
    "    display(df_partial)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 讀入並合併多個 CSV 檔案成單一 DataFrame"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "很多時候因為企業內部 ETL 或是數據處理的方式（比方說[利用 Airflow 處理批次數據](https://leemeng.tw/a-story-about-airflow-and-data-engineering-using-how-to-use-python-to-catch-up-with-latest-comics-as-an-example.html)），相同類型的數據可能會被分成多個不同的 CSV 檔案儲存。\n",
    "\n",
    "假設在本地端 `dataset` 資料夾內有 2 個 CSV 檔案，分別儲存鐵達尼號上不同乘客的數據："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "#ignore\n",
    "df = pd.read_csv(\"dataset/titanic-train.csv\")\n",
    "columns = list(df.columns)\n",
    "# columns = [c for c in columns if c != 'Name']\n",
    "df.Name = df.Name.apply(lambda x: x.split()[0].replace(',', ''))\n",
    "\n",
    "df.loc[4:6, columns].to_csv(\"dataset/passenger1.csv\", index=False)\n",
    "df.loc[12:14, columns].to_csv(\"dataset/passenger2.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Allen</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",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Moran</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>330877</td>\n",
       "      <td>8.4583</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>McCarthy</td>\n",
       "      <td>male</td>\n",
       "      <td>54.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>17463</td>\n",
       "      <td>51.8625</td>\n",
       "      <td>E46</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived  Pclass      Name   Sex   Age  SibSp  Parch  Ticket  \\\n",
       "0            5         0       3     Allen  male  35.0      0      0  373450   \n",
       "1            6         0       3     Moran  male   NaN      0      0  330877   \n",
       "2            7         0       1  McCarthy  male  54.0      0      0   17463   \n",
       "\n",
       "      Fare Cabin Embarked  \n",
       "0   8.0500   NaN        S  \n",
       "1   8.4583   NaN        Q  \n",
       "2  51.8625   E46        S  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.read_csv(\"dataset/passenger1.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "另外一個 CSV 內容："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>13</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Saundercock</td>\n",
       "      <td>male</td>\n",
       "      <td>20.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5. 2151</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>14</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Andersson</td>\n",
       "      <td>male</td>\n",
       "      <td>39.0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>347082</td>\n",
       "      <td>31.2750</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>15</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Vestrom</td>\n",
       "      <td>female</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>350406</td>\n",
       "      <td>7.8542</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived  Pclass         Name     Sex   Age  SibSp  Parch  \\\n",
       "0           13         0       3  Saundercock    male  20.0      0      0   \n",
       "1           14         0       3    Andersson    male  39.0      1      5   \n",
       "2           15         0       3      Vestrom  female  14.0      0      0   \n",
       "\n",
       "      Ticket     Fare  Cabin Embarked  \n",
       "0  A/5. 2151   8.0500    NaN        S  \n",
       "1     347082  31.2750    NaN        S  \n",
       "2     350406   7.8542    NaN        S  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.read_csv(\"dataset/passenger2.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "注意上面 2 個 DataFrames 的內容雖然分別代表不同乘客，其格式卻是一模一樣。這種時候你可以使用 `pd.concat` 將分散在不同 CSV 的乘客數據合併成單一 DataFrame，方便之後處理："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Allen</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",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Moran</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>330877</td>\n",
       "      <td>8.4583</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>McCarthy</td>\n",
       "      <td>male</td>\n",
       "      <td>54.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>17463</td>\n",
       "      <td>51.8625</td>\n",
       "      <td>E46</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>13</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Saundercock</td>\n",
       "      <td>male</td>\n",
       "      <td>20.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5. 2151</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>14</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Andersson</td>\n",
       "      <td>male</td>\n",
       "      <td>39.0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>347082</td>\n",
       "      <td>31.2750</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>15</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Vestrom</td>\n",
       "      <td>female</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>350406</td>\n",
       "      <td>7.8542</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived  Pclass         Name     Sex   Age  SibSp  Parch  \\\n",
       "0            5         0       3        Allen    male  35.0      0      0   \n",
       "1            6         0       3        Moran    male   NaN      0      0   \n",
       "2            7         0       1     McCarthy    male  54.0      0      0   \n",
       "3           13         0       3  Saundercock    male  20.0      0      0   \n",
       "4           14         0       3    Andersson    male  39.0      1      5   \n",
       "5           15         0       3      Vestrom  female  14.0      0      0   \n",
       "\n",
       "      Ticket     Fare Cabin Embarked  \n",
       "0     373450   8.0500   NaN        S  \n",
       "1     330877   8.4583   NaN        Q  \n",
       "2      17463  51.8625   E46        S  \n",
       "3  A/5. 2151   8.0500   NaN        S  \n",
       "4     347082  31.2750   NaN        S  \n",
       "5     350406   7.8542   NaN        S  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from glob import glob\n",
    "files = glob(\"dataset/passenger*.csv\")\n",
    "\n",
    "df = pd.concat([pd.read_csv(f) for f in files])\n",
    "df.reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你還可以使用 `reset_index` 函式來重置串接後的 DataFrame 索引。\n",
    "\n",
    "前面說過很多 pandas 函式**預設**的 `axis` 參數為 `0`，代表著以**列（row）**為單位做特定的操作。在 `pd.concat` 的例子中則是將 2 個同樣格式的 DataFrames 依照**列**串接起來。\n",
    "\n",
    "有時候同一筆數據的不同特徵值（features）會被存在不同檔案裡頭。以鐵達尼號的數據集舉例："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "#ignore\n",
    "df = pd.read_csv(\"dataset/titanic-train.csv\")\n",
    "df.Name = df.Name.apply(lambda x: x.split()[0].replace(',', ''))\n",
    "df.iloc[:, :4].head().to_csv(\"dataset/feature_set1.csv\", index=False)\n",
    "df.iloc[:, 4:].head().to_csv(\"dataset/feature_set2.csv\", index=False)"
   ]
  },
  {
   "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>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Name</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</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Cumings</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Heikkinen</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Futrelle</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Allen</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived  Pclass       Name\n",
       "0            1         0       3     Braund\n",
       "1            2         1       1    Cumings\n",
       "2            3         1       3  Heikkinen\n",
       "3            4         1       1   Futrelle\n",
       "4            5         0       3      Allen"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.read_csv(\"dataset/feature_set1.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "除了乘客名稱以外，其他如年齡以及性別等特徵值則被存在另個 CSV 裡頭："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>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": [
       "      Sex   Age  SibSp  Parch            Ticket     Fare Cabin Embarked\n",
       "0    male  22.0      1      0         A/5 21171   7.2500   NaN        S\n",
       "1  female  38.0      1      0          PC 17599  71.2833   C85        C\n",
       "2  female  26.0      0      0  STON/O2. 3101282   7.9250   NaN        S\n",
       "3  female  35.0      1      0            113803  53.1000  C123        S\n",
       "4    male  35.0      0      0            373450   8.0500   NaN        S"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.read_csv(\"dataset/feature_set2.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "假設這 2 個 CSV 檔案裡頭**同列**對應到同個乘客，則你可以很輕鬆地用 `pd.concat` 函式搭配 `axis=1` 將不同 DataFrames 依照**行（column）**串接："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Braund</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Cumings</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Heikkinen</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Futrelle</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Allen</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       Name     Sex   Age  SibSp  Parch  \\\n",
       "0            1         0       3     Braund    male  22.0      1      0   \n",
       "1            2         1       1    Cumings  female  38.0      1      0   \n",
       "2            3         1       3  Heikkinen  female  26.0      0      0   \n",
       "3            4         1       1   Futrelle  female  35.0      1      0   \n",
       "4            5         0       3      Allen    male  35.0      0      0   \n",
       "\n",
       "             Ticket     Fare Cabin Embarked  \n",
       "0         A/5 21171   7.2500   NaN        S  \n",
       "1          PC 17599  71.2833   C85        C  \n",
       "2  STON/O2. 3101282   7.9250   NaN        S  \n",
       "3            113803  53.1000  C123        S  \n",
       "4            373450   8.0500   NaN        S  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "files = glob(\"dataset/feature_set*.csv\")\n",
    "pd.concat([pd.read_csv(f) for f in files], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## 客製化 DataFrame 顯示設定"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "雖然 pandas 會盡可能地將一個 DataFrame 完整且漂亮地呈現出來，有時候你還是會想要改變預設的顯示方式。這節列出一些常見的使用情境。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 完整顯示所有欄位"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "有時候一個 DataFrame 裡頭的欄位太多， pandas 會自動省略某些中間欄位以保持頁面整潔：\n",
    "\n",
    "```python\n",
    "df = pd.util.testing.makeCustomDataframe(5, 25)\n",
    "df\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!image\n",
    "- dark\n",
    "- pandas/long_dataframe.jpg"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "但如果你無論如何都想要顯示所有欄位以方便一次查看，可以透過 `pd.set_option` 函式來改變 `display.max_columns` 設定：\n",
    "\n",
    "```python\n",
    "pd.set_option(\"display.max_columns\", None)\n",
    "df\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!image\n",
    "- dark\n",
    "- pandas/long_full_dataframe.jpg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "#ignore\n",
    "pd.reset_option(\"display.max_columns\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "注意 `...` 消失了。另外你也可以使用 `T` 來轉置（transpose）當前 DataFrame，垂直顯示所有欄位：\n",
    "\n",
    "```python\n",
    "# 注意轉置後 `head(15)` 代表選擇前 15 個欄位\n",
    "df.T.head(15)\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!image\n",
    "- dark\n",
    "- pandas/transposed_dataframe.jpg"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "這個測試用的 DataFrame 欄位裡頭的 `C` 正代表著 column。你可以在 [pandas 官方文件裡查看其他常用的顯示設定](https://pandas.pydata.org/pandas-docs/stable/user_guide/options.html#frequently-used-options)。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 減少顯示的欄位長度"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "這邊你一樣可以透過 `pd.set_option` 函式來限制鐵達尼號資料集裡頭 `Name` 欄位的顯示長度："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "display.max_colwidth 預設值： 50\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Braund, Mr. Owen Harris</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Cumings, Mrs. John Bradley (Florence Briggs Th...</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Heikkinen, Miss. Laina</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </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",
       "\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",
       "\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  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "注意 Name 欄位的長度被改變了：\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Braund...</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Cuming...</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Heikki...</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O...</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived  Pclass       Name     Sex   Age  SibSp  Parch  \\\n",
       "0          1           0       3  Braund...    male  22.0      1      0   \n",
       "1          2           1       1  Cuming...  female  38.0      1      0   \n",
       "2          3           1       3  Heikki...  female  26.0      0      0   \n",
       "\n",
       "      Ticket     Fare Cabin Embarked  \n",
       "0  A/5 21171   7.2500   NaN        S  \n",
       "1   PC 17599  71.2833   C85        C  \n",
       "2  STON/O...   7.9250   NaN        S  "
      ]
     },
     "execution_count": 135,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import display\n",
    "print(\"display.max_colwidth 預設值：\", \n",
    "      pd.get_option(\"display.max_colwidth\"))\n",
    "\n",
    "# 使用預設設定來顯示 DataFrame\n",
    "df = pd.read_csv('http://bit.ly/kaggletrain')\n",
    "display(df.head(3))\n",
    "\n",
    "print(\"注意 Name 欄位的長度被改變了：\")\n",
    "# 客製化顯示（global）\n",
    "pd.set_option(\"display.max_colwidth\", 10)\n",
    "df.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 改變浮點數顯示位數"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "除了欄位長度以外，你常常會想要改變浮點數（float）顯示的小數點位數："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Braund...</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>7.2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Cuming...</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>71.3</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Heikki...</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O...</td>\n",
       "      <td>7.9</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived  Pclass       Name     Sex   Age  SibSp  Parch  \\\n",
       "0          1           0       3  Braund...    male  22.0      1      0   \n",
       "1          2           1       1  Cuming...  female  38.0      1      0   \n",
       "2          3           1       3  Heikki...  female  26.0      0      0   \n",
       "\n",
       "      Ticket  Fare Cabin Embarked  \n",
       "0  A/5 21171   7.2   NaN        S  \n",
       "1   PC 17599  71.3   C85        C  \n",
       "2  STON/O...   7.9   NaN        S  "
      ]
     },
     "execution_count": 132,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option(\"display.precision\", 1)\n",
    "df.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你會發現 `Fare` 欄位現在只顯示小數點**後一位**的數值了。另外注意剛剛設定的 `max_colwidth` 是會被套用到所有 DataFrame 的。因此這個 DataFrame 的 `Name` 欄位顯示的寬度還跟上個 DataFrame 相同：都被縮減了。\n",
    "\n",
    "想要將所有調整過的設定初始化，可以執行：\n",
    "\n",
    "```python\n",
    "pd.reset_option(\"all\")\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [],
   "source": [
    "#ignore\n",
    "from IPython.display import clear_output\n",
    "pd.reset_option(\"all\")\n",
    "clear_output()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "其他常用的 options 包含：\n",
    "- `max_rows`\n",
    "- `max_columns`\n",
    "- `date_yearfirst` \n",
    "\n",
    "等等。執行 `pd.describe_option()` 可以顯示所有可供使用的 options，但如果你是在 Jupyter 筆記本內使用 pandas 的話，我推薦直接在 `set_option` 函式的括號裡輸入 `Shift` + `tab` 顯示所有選項："
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!mp4\n",
    "- dark\n",
    "- images/pandas/pd_set_option.mp4"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 為特定 DataFrame 加點樣式"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`pd.set_option` 函式在你想要把某些顯示設定套用到**所有** DataFrames 時很好用。不過很多時候你會想要讓不同 DataFrame 有不同的顯示設定或樣式（styling）。\n",
    "\n",
    "比方說針對下面這個只有 10 筆數據的 DataFrame，你想要跟上一節一樣把 `Fare` 欄位弄成只有小數點後一位，但又不想影響到其他 DataFrame 或是其他欄位："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "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>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</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",
       "    <tr>\n",
       "      <th>182</th>\n",
       "      <td>183</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>9.0</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>347077</td>\n",
       "      <td>31.3875</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>376</th>\n",
       "      <td>377</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>22.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>C 7077</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>659</th>\n",
       "      <td>660</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>58.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>35273</td>\n",
       "      <td>113.2750</td>\n",
       "      <td>D48</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>431</th>\n",
       "      <td>432</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>29.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>376564</td>\n",
       "      <td>16.1000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>539</th>\n",
       "      <td>540</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>22.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>13568</td>\n",
       "      <td>49.5000</td>\n",
       "      <td>B39</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>478</th>\n",
       "      <td>479</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>350060</td>\n",
       "      <td>7.5208</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>791</th>\n",
       "      <td>792</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>male</td>\n",
       "      <td>16.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>239865</td>\n",
       "      <td>26.0000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>303</th>\n",
       "      <td>304</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>female</td>\n",
       "      <td>29.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>226593</td>\n",
       "      <td>12.3500</td>\n",
       "      <td>E101</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>71</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>male</td>\n",
       "      <td>32.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>C.A. 33111</td>\n",
       "      <td>10.5000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     PassengerId  Survived  Pclass     Sex   Age  SibSp  Parch      Ticket  \\\n",
       "4              5         0       3    male  35.0      0      0      373450   \n",
       "182          183         0       3    male   9.0      4      2      347077   \n",
       "376          377         1       3  female  22.0      0      0      C 7077   \n",
       "659          660         0       1    male  58.0      0      2       35273   \n",
       "431          432         1       3  female  29.0      1      0      376564   \n",
       "539          540         1       1  female  22.0      0      2       13568   \n",
       "478          479         0       3    male  22.0      0      0      350060   \n",
       "791          792         0       2    male  16.0      0      0      239865   \n",
       "303          304         1       2  female  29.0      0      0      226593   \n",
       "70            71         0       2    male  32.0      0      0  C.A. 33111   \n",
       "\n",
       "         Fare Cabin Embarked  \n",
       "4      8.0500   NaN        S  \n",
       "182   31.3875   NaN        S  \n",
       "376    7.2500   NaN        S  \n",
       "659  113.2750   D48        C  \n",
       "431   16.1000   NaN        S  \n",
       "539   49.5000   B39        C  \n",
       "478    7.5208   NaN        S  \n",
       "791   26.0000   NaN        S  \n",
       "303   12.3500  E101        Q  \n",
       "70    10.5000   NaN        S  "
      ]
     },
     "execution_count": 172,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 隨機抽樣 10 筆數據來做 styling\n",
    "df_sample = df.sample(n=10, random_state=9527).drop('Name', axis=1)\n",
    "df_sample.Age.fillna(int(df.Age.mean()), inplace=True)\n",
    "df_sample"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "這時候你可以使用 [pandas Styler](https://pandas.pydata.org/pandas-docs/stable/user_guide/style.html) 底下的 `format` 函式來做到這件事情：\n",
    "\n",
    "```python\n",
    "# 一個典型 chain pandas 函式的例子\n",
    "(df_sample.style\n",
    "     .format('{:.1f}', subset='Fare')\n",
    "     .set_caption('★五顏六色の鐵達尼號數據集☆')\n",
    "     .hide_index()\n",
    "     .bar('Age', vmin=0)\n",
    "     .highlight_max('Survived')\n",
    "     .background_gradient('Greens', \n",
    "                          subset='Fare')\n",
    "     .highlight_null()\n",
    ")\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!image\n",
    "- pandas/pandas_styling.jpg"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "如果你從來沒有用過 [df.style](https://pandas.pydata.org/pandas-docs/stable/user_guide/style.html)，這應該是你這輩子看過最繽紛的 DataFrame。\n",
    "\n",
    "從上而下，上述程式碼對此 DataFrame 做了以下 styling：\n",
    "- 將 `Fare` 欄位的數值顯示限制到小數後第一位\n",
    "- 添加一個標題輔助說明\n",
    "- 隱藏索引（注意最左邊！）\n",
    "- 將 `Age` 欄位依數值大小畫條狀圖\n",
    "- 將 `Survived` 最大的值 highlight\n",
    "- 將 `Fare` 欄位依數值畫綠色的 colormap\n",
    "- 將整個 DataFrame 的空值顯示為紅色\n",
    "\n",
    "`pd.DataFrame.style` 會回傳一個 Styler。你已經看到除了 `format` 函式以外，還有很多其他函式可以讓你為 DataFrame 添加樣式。使用 `format` 函式的最大好處是你不需要用像是 `round` 等函式去修改**實際**數值，而只是改變**呈現結果**而已。\n",
    "\n",
    "熟悉 styling 技巧能讓你不需畫圖就能輕鬆與他人分享簡單的分析結果，也能凸顯你想讓他們關注的事物。小提醒：為了讓你能一次掌握常用函式，我把能加的樣式都加了。實務上你應該思考[什麼視覺變數是必要的](https://leemeng.tw/data-visualization-from-matplotlib-to-ggplot2.html)，而不是盲目地添加樣式。\n",
    "\n",
    "另外值得一提的是 pandas 函式都會回傳處理後的結果，而不是直接修改原始 DataFrame。這讓你可以輕鬆地把多個函式串（chain）成一個複雜的數據處理 pipeline，但又不會影響到最原始的數據："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "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>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</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",
       "    <tr>\n",
       "      <th>182</th>\n",
       "      <td>183</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>9.0</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>347077</td>\n",
       "      <td>31.3875</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>376</th>\n",
       "      <td>377</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>22.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>C 7077</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>659</th>\n",
       "      <td>660</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>58.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>35273</td>\n",
       "      <td>113.2750</td>\n",
       "      <td>D48</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>431</th>\n",
       "      <td>432</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>29.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>376564</td>\n",
       "      <td>16.1000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>539</th>\n",
       "      <td>540</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>22.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>13568</td>\n",
       "      <td>49.5000</td>\n",
       "      <td>B39</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>478</th>\n",
       "      <td>479</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>350060</td>\n",
       "      <td>7.5208</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>791</th>\n",
       "      <td>792</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>male</td>\n",
       "      <td>16.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>239865</td>\n",
       "      <td>26.0000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>303</th>\n",
       "      <td>304</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>female</td>\n",
       "      <td>29.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>226593</td>\n",
       "      <td>12.3500</td>\n",
       "      <td>E101</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>71</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>male</td>\n",
       "      <td>32.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>C.A. 33111</td>\n",
       "      <td>10.5000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     PassengerId  Survived  Pclass     Sex   Age  SibSp  Parch      Ticket  \\\n",
       "4              5         0       3    male  35.0      0      0      373450   \n",
       "182          183         0       3    male   9.0      4      2      347077   \n",
       "376          377         1       3  female  22.0      0      0      C 7077   \n",
       "659          660         0       1    male  58.0      0      2       35273   \n",
       "431          432         1       3  female  29.0      1      0      376564   \n",
       "539          540         1       1  female  22.0      0      2       13568   \n",
       "478          479         0       3    male  22.0      0      0      350060   \n",
       "791          792         0       2    male  16.0      0      0      239865   \n",
       "303          304         1       2  female  29.0      0      0      226593   \n",
       "70            71         0       2    male  32.0      0      0  C.A. 33111   \n",
       "\n",
       "         Fare Cabin Embarked  \n",
       "4      8.0500   NaN        S  \n",
       "182   31.3875   NaN        S  \n",
       "376    7.2500   NaN        S  \n",
       "659  113.2750   D48        C  \n",
       "431   16.1000   NaN        S  \n",
       "539   49.5000   B39        C  \n",
       "478    7.5208   NaN        S  \n",
       "791   26.0000   NaN        S  \n",
       "303   12.3500  E101        Q  \n",
       "70    10.5000   NaN        S  "
      ]
     },
     "execution_count": 173,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_sample"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "瞧！原來的 DataFrame 還是挺淳樸的。注意 `Fare` 欄位裡的小數點並沒有因為剛剛的 styling 而變少，而這讓你在呈現 DataFrame 時有最大的彈性。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## 數據清理 & 整理"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "這節列出一些十分常用的數據清理與整理技巧，如處理空值（null value）以及切割欄位。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 處理空值"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "世界總是殘酷，很多時候手上的 DataFrame 裡頭會有不存在的值，如底下一格格額外顯眼的 `NaN`："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6bAs9RrtPf</th>\n",
       "      <td>-1.864693</td>\n",
       "      <td>0.395997</td>\n",
       "      <td>-1.057175</td>\n",
       "      <td>1.261987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>P4pLVw54w1</th>\n",
       "      <td>0.817813</td>\n",
       "      <td>-0.893529</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KkhqmjKghu</th>\n",
       "      <td>0.182929</td>\n",
       "      <td>0.892841</td>\n",
       "      <td>1.487175</td>\n",
       "      <td>-1.218580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S4ppq42BKt</th>\n",
       "      <td>1.113693</td>\n",
       "      <td>0.979332</td>\n",
       "      <td>-0.761886</td>\n",
       "      <td>0.026489</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mu2ryPVR0x</th>\n",
       "      <td>0.387857</td>\n",
       "      <td>0.334798</td>\n",
       "      <td>1.321586</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "6bAs9RrtPf -1.864693  0.395997 -1.057175  1.261987\n",
       "P4pLVw54w1  0.817813 -0.893529       NaN       NaN\n",
       "KkhqmjKghu  0.182929  0.892841  1.487175 -1.218580\n",
       "S4ppq42BKt  1.113693  0.979332 -0.761886  0.026489\n",
       "Mu2ryPVR0x  0.387857  0.334798  1.321586       NaN"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.util.testing.makeMissingDataframe().head()\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你可以利用 `fillna` 函式將 DataFrame 裡頭所有不存在的值設為 `0`："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6bAs9RrtPf</th>\n",
       "      <td>-1.864693</td>\n",
       "      <td>0.395997</td>\n",
       "      <td>-1.057175</td>\n",
       "      <td>1.261987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>P4pLVw54w1</th>\n",
       "      <td>0.817813</td>\n",
       "      <td>-0.893529</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KkhqmjKghu</th>\n",
       "      <td>0.182929</td>\n",
       "      <td>0.892841</td>\n",
       "      <td>1.487175</td>\n",
       "      <td>-1.218580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S4ppq42BKt</th>\n",
       "      <td>1.113693</td>\n",
       "      <td>0.979332</td>\n",
       "      <td>-0.761886</td>\n",
       "      <td>0.026489</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mu2ryPVR0x</th>\n",
       "      <td>0.387857</td>\n",
       "      <td>0.334798</td>\n",
       "      <td>1.321586</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "6bAs9RrtPf -1.864693  0.395997 -1.057175  1.261987\n",
       "P4pLVw54w1  0.817813 -0.893529  0.000000  0.000000\n",
       "KkhqmjKghu  0.182929  0.892841  1.487175 -1.218580\n",
       "S4ppq42BKt  1.113693  0.979332 -0.761886  0.026489\n",
       "Mu2ryPVR0x  0.387857  0.334798  1.321586  0.000000"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.fillna(0) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "當然，這個操作的前提是你確定在當前分析的情境下，將不存在的值視為 `0` 這件事情是沒有問題的。\n",
    "\n",
    "針對字串欄位，你也可以將空值設定成任何容易識別的值，讓自己及他人明確了解此 DataFrame 的數據品質："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>C0</th>\n",
       "      <th>C_l0_g0</th>\n",
       "      <th>C_l0_g1</th>\n",
       "      <th>C_l0_g2</th>\n",
       "      <th>C_l0_g3</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R0</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>R_l0_g0</th>\n",
       "      <td>R0C0</td>\n",
       "      <td>R0C1</td>\n",
       "      <td>R0C2</td>\n",
       "      <td>R0C3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R_l0_g1</th>\n",
       "      <td>R1C0</td>\n",
       "      <td>R1C1</td>\n",
       "      <td>R1C2</td>\n",
       "      <td>R1C3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R_l0_g2</th>\n",
       "      <td>R2C0</td>\n",
       "      <td>R2C1</td>\n",
       "      <td>R2C2</td>\n",
       "      <td>R2C3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R_l0_g3</th>\n",
       "      <td>R3C0</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>R3C2</td>\n",
       "      <td>R3C3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>R_l0_g4</th>\n",
       "      <td>Unknown</td>\n",
       "      <td>R4C1</td>\n",
       "      <td>R4C2</td>\n",
       "      <td>R4C3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "C0       C_l0_g0  C_l0_g1 C_l0_g2 C_l0_g3\n",
       "R0                                       \n",
       "R_l0_g0     R0C0     R0C1    R0C2    R0C3\n",
       "R_l0_g1     R1C0     R1C1    R1C2    R1C3\n",
       "R_l0_g2     R2C0     R2C1    R2C2    R2C3\n",
       "R_l0_g3     R3C0  Unknown    R3C2    R3C3\n",
       "R_l0_g4  Unknown     R4C1    R4C2    R4C3"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.util.testing.makeMissingCustomDataframe(nrows=5, \n",
    "                                                ncols=4, \n",
    "                                                dtype=str)\n",
    "df.fillna(\"Unknown\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 捨棄不需要的行列"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "給定一個初始 DataFrame："
   ]
  },
  {
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>87dl3rTYXR</th>\n",
       "      <td>0.048544</td>\n",
       "      <td>0.133999</td>\n",
       "      <td>0.885555</td>\n",
       "      <td>1.164883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DE0ptDHz2Q</th>\n",
       "      <td>0.551378</td>\n",
       "      <td>0.012621</td>\n",
       "      <td>0.256241</td>\n",
       "      <td>0.197140</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>jiWMAjniDH</th>\n",
       "      <td>-1.275694</td>\n",
       "      <td>-1.473950</td>\n",
       "      <td>-1.602433</td>\n",
       "      <td>-0.286107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45lFb2QT9g</th>\n",
       "      <td>1.348614</td>\n",
       "      <td>2.235074</td>\n",
       "      <td>1.471781</td>\n",
       "      <td>0.449909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OOOtaUhOOp</th>\n",
       "      <td>-0.297613</td>\n",
       "      <td>-1.332934</td>\n",
       "      <td>0.194135</td>\n",
       "      <td>0.654267</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "87dl3rTYXR  0.048544  0.133999  0.885555  1.164883\n",
       "DE0ptDHz2Q  0.551378  0.012621  0.256241  0.197140\n",
       "jiWMAjniDH -1.275694 -1.473950 -1.602433 -0.286107\n",
       "45lFb2QT9g  1.348614  2.235074  1.471781  0.449909\n",
       "OOOtaUhOOp -0.297613 -1.332934  0.194135  0.654267"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.util.testing.makeDataFrame().head()\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你可以使用 `drop` 函式來捨棄不需要的欄位。記得將 `axis` 設為 1："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>A</th>\n",
       "      <th>C</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>87dl3rTYXR</th>\n",
       "      <td>0.048544</td>\n",
       "      <td>0.885555</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DE0ptDHz2Q</th>\n",
       "      <td>0.551378</td>\n",
       "      <td>0.256241</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>jiWMAjniDH</th>\n",
       "      <td>-1.275694</td>\n",
       "      <td>-1.602433</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45lFb2QT9g</th>\n",
       "      <td>1.348614</td>\n",
       "      <td>1.471781</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OOOtaUhOOp</th>\n",
       "      <td>-0.297613</td>\n",
       "      <td>0.194135</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         C\n",
       "87dl3rTYXR  0.048544  0.885555\n",
       "DE0ptDHz2Q  0.551378  0.256241\n",
       "jiWMAjniDH -1.275694 -1.602433\n",
       "45lFb2QT9g  1.348614  1.471781\n",
       "OOOtaUhOOp -0.297613  0.194135"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "columns = ['B', 'D']\n",
    "df.drop(columns, axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "同理，你也可以捨棄特定列（row）："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>87dl3rTYXR</th>\n",
       "      <td>0.048544</td>\n",
       "      <td>0.133999</td>\n",
       "      <td>0.885555</td>\n",
       "      <td>1.164883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DE0ptDHz2Q</th>\n",
       "      <td>0.551378</td>\n",
       "      <td>0.012621</td>\n",
       "      <td>0.256241</td>\n",
       "      <td>0.197140</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>jiWMAjniDH</th>\n",
       "      <td>-1.275694</td>\n",
       "      <td>-1.473950</td>\n",
       "      <td>-1.602433</td>\n",
       "      <td>-0.286107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45lFb2QT9g</th>\n",
       "      <td>1.348614</td>\n",
       "      <td>2.235074</td>\n",
       "      <td>1.471781</td>\n",
       "      <td>0.449909</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "87dl3rTYXR  0.048544  0.133999  0.885555  1.164883\n",
       "DE0ptDHz2Q  0.551378  0.012621  0.256241  0.197140\n",
       "jiWMAjniDH -1.275694 -1.473950 -1.602433 -0.286107\n",
       "45lFb2QT9g  1.348614  2.235074  1.471781  0.449909"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.drop('OOOtaUhOOp')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 重置並捨棄索引"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "很多時候你會想要重置一個 DataFrame 的索引，以方便使用 `loc` 或 `iloc` **屬性**來存取想要的數據。\n",
    "\n",
    "給定一個 DataFrame："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>LI5Y0d3Ygk</th>\n",
       "      <td>0.704950</td>\n",
       "      <td>0.610508</td>\n",
       "      <td>-1.686467</td>\n",
       "      <td>0.165393</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9EkrfcGaKh</th>\n",
       "      <td>1.088294</td>\n",
       "      <td>-0.947646</td>\n",
       "      <td>1.382226</td>\n",
       "      <td>-1.410468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Jf891pQjSh</th>\n",
       "      <td>0.137240</td>\n",
       "      <td>1.490714</td>\n",
       "      <td>1.236335</td>\n",
       "      <td>-0.270623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Q9O2vu4Mg1</th>\n",
       "      <td>-0.375816</td>\n",
       "      <td>2.308924</td>\n",
       "      <td>-1.735557</td>\n",
       "      <td>1.734371</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>z0xoUxOAWW</th>\n",
       "      <td>1.067996</td>\n",
       "      <td>0.474090</td>\n",
       "      <td>-0.492757</td>\n",
       "      <td>-0.042121</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "LI5Y0d3Ygk  0.704950  0.610508 -1.686467  0.165393\n",
       "9EkrfcGaKh  1.088294 -0.947646  1.382226 -1.410468\n",
       "Jf891pQjSh  0.137240  1.490714  1.236335 -0.270623\n",
       "Q9O2vu4Mg1 -0.375816  2.308924 -1.735557  1.734371\n",
       "z0xoUxOAWW  1.067996  0.474090 -0.492757 -0.042121"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.util.testing.makeDataFrame().head()\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你可以使用 `reset_index` 函式來重置此 DataFrame 的索引並輕鬆存取想要的部分："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "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>index</th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>LI5Y0d3Ygk</td>\n",
       "      <td>0.704950</td>\n",
       "      <td>0.610508</td>\n",
       "      <td>-1.686467</td>\n",
       "      <td>0.165393</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9EkrfcGaKh</td>\n",
       "      <td>1.088294</td>\n",
       "      <td>-0.947646</td>\n",
       "      <td>1.382226</td>\n",
       "      <td>-1.410468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Jf891pQjSh</td>\n",
       "      <td>0.137240</td>\n",
       "      <td>1.490714</td>\n",
       "      <td>1.236335</td>\n",
       "      <td>-0.270623</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        index         A         B         C         D\n",
       "0  LI5Y0d3Ygk  0.704950  0.610508 -1.686467  0.165393\n",
       "1  9EkrfcGaKh  1.088294 -0.947646  1.382226 -1.410468\n",
       "2  Jf891pQjSh  0.137240  1.490714  1.236335 -0.270623"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.reset_index(inplace=True)\n",
    "df.iloc[:3, :]\n",
    "# 豆知識：因為 iloc 是屬性而非函式，\n",
    "# 因此你得使用 [] 而非 () 存取數據"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "將函式的 `inplace` 參數設為 `True` 會讓 pandas 直接修改 `df`。一般來說 pandas 裡的函式並不會修改原始 DataFrame，這樣可以保證原始數據不會受到任何函式的影響。\n",
    "\n",
    "當你不想要原來的 DataFrame `df` 受到 `reset_index` 函式的影響，則可以將處理後的結果交給一個新 DataFrame（比方說 `df1`）："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>fNMRqfPnw1</th>\n",
       "      <td>0.458255</td>\n",
       "      <td>1.736118</td>\n",
       "      <td>-1.659602</td>\n",
       "      <td>2.325742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>xeIJ4tJuxW</th>\n",
       "      <td>-1.102030</td>\n",
       "      <td>1.321306</td>\n",
       "      <td>-0.262345</td>\n",
       "      <td>0.864090</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RDAJlNsfDS</th>\n",
       "      <td>1.047750</td>\n",
       "      <td>-0.420285</td>\n",
       "      <td>0.757464</td>\n",
       "      <td>0.384514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>kIOVU4EU79</th>\n",
       "      <td>-0.843373</td>\n",
       "      <td>0.235789</td>\n",
       "      <td>-1.211142</td>\n",
       "      <td>0.656130</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OKuRXq4eEK</th>\n",
       "      <td>-1.340040</td>\n",
       "      <td>0.674713</td>\n",
       "      <td>0.768126</td>\n",
       "      <td>-0.500718</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "fNMRqfPnw1  0.458255  1.736118 -1.659602  2.325742\n",
       "xeIJ4tJuxW -1.102030  1.321306 -0.262345  0.864090\n",
       "RDAJlNsfDS  1.047750 -0.420285  0.757464  0.384514\n",
       "kIOVU4EU79 -0.843373  0.235789 -1.211142  0.656130\n",
       "OKuRXq4eEK -1.340040  0.674713  0.768126 -0.500718"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.458255</td>\n",
       "      <td>1.736118</td>\n",
       "      <td>-1.659602</td>\n",
       "      <td>2.325742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.102030</td>\n",
       "      <td>1.321306</td>\n",
       "      <td>-0.262345</td>\n",
       "      <td>0.864090</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.047750</td>\n",
       "      <td>-0.420285</td>\n",
       "      <td>0.757464</td>\n",
       "      <td>0.384514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.843373</td>\n",
       "      <td>0.235789</td>\n",
       "      <td>-1.211142</td>\n",
       "      <td>0.656130</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.340040</td>\n",
       "      <td>0.674713</td>\n",
       "      <td>0.768126</td>\n",
       "      <td>-0.500718</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          A         B         C         D\n",
       "0  0.458255  1.736118 -1.659602  2.325742\n",
       "1 -1.102030  1.321306 -0.262345  0.864090\n",
       "2  1.047750 -0.420285  0.757464  0.384514\n",
       "3 -0.843373  0.235789 -1.211142  0.656130\n",
       "4 -1.340040  0.674713  0.768126 -0.500718"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df = pd.util.testing.makeDataFrame().head()\n",
    "df1 = df.reset_index(drop=True)\n",
    "display(df)\n",
    "display(df1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "透過這樣的方式，pandas 讓你可以放心地對原始數據做任何壞壞的事情而不會產生任何不好的影響。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 將字串切割成多個欄位"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在處理文本數據時，很多時候你會想要把一個字串欄位拆成多個欄位以方便後續處理。\n",
    "\n",
    "給定一個簡單 DataFrame："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>feature</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>大雄</td>\n",
       "      <td>膽小, 翻花繩</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>胖虎</td>\n",
       "      <td>粗魯, 演唱會</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  name  feature\n",
       "0   大雄  膽小, 翻花繩\n",
       "1   胖虎  粗魯, 演唱會"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame({\n",
    "    \"name\": [\"大雄\", \"胖虎\"], \n",
    "    \"feature\": [\"膽小, 翻花繩\", \"粗魯, 演唱會\"]\n",
    "})\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你可能會想把這個 DataFrame 的 `feature` 欄位分成不同欄位，這時候利用 `str` 將字串取出，並透過 `expand=True` 將字串切割的結果擴大成（expand）成一個 DataFrame："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "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>name</th>\n",
       "      <th>feature</th>\n",
       "      <th>性格</th>\n",
       "      <th>特技</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>大雄</td>\n",
       "      <td>膽小, 翻花繩</td>\n",
       "      <td>膽小</td>\n",
       "      <td>翻花繩</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>胖虎</td>\n",
       "      <td>粗魯, 演唱會</td>\n",
       "      <td>粗魯</td>\n",
       "      <td>演唱會</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  name  feature  性格    特技\n",
       "0   大雄  膽小, 翻花繩  膽小   翻花繩\n",
       "1   胖虎  粗魯, 演唱會  粗魯   演唱會"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[['性格', '特技']] = df.feature.str.split(',', expand=True)\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "注意我們使用 `df[columns] = ...` 的形式將字串切割出來的 2 個新欄位分別指定成 `性格` 與 `特技`。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 將 list 分成多個欄位"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "有時候一個欄位裡頭的值為 Python `list`："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "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>name</th>\n",
       "      <th>feature</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>大雄</td>\n",
       "      <td>[膽小, 翻花繩]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>胖虎</td>\n",
       "      <td>[粗魯, 演唱會]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  name    feature\n",
       "0   大雄  [膽小, 翻花繩]\n",
       "1   胖虎  [粗魯, 演唱會]"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame({\n",
    "    \"name\": [\"大雄\", \"胖虎\"], \n",
    "    \"feature\": [[\"膽小\", \"翻花繩\"], [\"粗魯\", \"演唱會\"]]\n",
    "})\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "這時則可以使用 `tolist` 函式做到跟剛剛字串切割相同的效果："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "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>性格</th>\n",
       "      <th>特技</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>膽小</td>\n",
       "      <td>翻花繩</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>粗魯</td>\n",
       "      <td>演唱會</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   性格   特技\n",
       "0  膽小  翻花繩\n",
       "1  粗魯  演唱會"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols = ['性格', '特技']\n",
    "pd.DataFrame(df.feature.tolist(), columns=cols)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你也可以使用 `apply(pd.Series)` 的方式達到一樣的效果："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>膽小</td>\n",
       "      <td>翻花繩</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>粗魯</td>\n",
       "      <td>演唱會</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    0    1\n",
       "0  膽小  翻花繩\n",
       "1  粗魯  演唱會"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.feature.apply(pd.Series)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "遇到以 Python `list` 呈現欄位數據的情境不少，這些函式能讓你少抓點頭。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## 取得想要關注的數據"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "通常你會需要依照各種不同的分析情境，將整個 DataFrame 裡頭的一部份數據取出並進一步分析。這節內容讓你能夠輕鬆取得想要關注的數據。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 基本數據切割"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在 pandas 裡頭，切割（Slice）DataFrame 裡頭一部份數據出來做分析是稀鬆平常的事情。讓我們再次以鐵達尼號數據集為例："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 272,
   "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>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>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     Sex   Age  SibSp  Parch  \\\n",
       "0            1         0       3    male  22.0      1      0   \n",
       "1            2         1       1  female  38.0      1      0   \n",
       "2            3         1       3  female  26.0      0      0   \n",
       "3            4         1       1  female  35.0      1      0   \n",
       "4            5         0       3    male  35.0      0      0   \n",
       "\n",
       "             Ticket     Fare Cabin Embarked  \n",
       "0         A/5 21171   7.2500   NaN        S  \n",
       "1          PC 17599  71.2833   C85        C  \n",
       "2  STON/O2. 3101282   7.9250   NaN        S  \n",
       "3            113803  53.1000  C123        S  \n",
       "4            373450   8.0500   NaN        S  "
      ]
     },
     "execution_count": 272,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('http://bit.ly/kaggletrain')\n",
    "df = df.drop(\"Name\", axis=1)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你可以透過 `loc` 以及 `:` 的方式輕鬆選取從某個起始欄位 C1 到結束欄位 C2 的所有欄位，而無需將中間的欄位一一列出："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 273,
   "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>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Pclass     Sex   Age  SibSp  Parch            Ticket\n",
       "0       3    male  22.0      1      0         A/5 21171\n",
       "1       1  female  38.0      1      0          PC 17599\n",
       "2       3  female  26.0      0      0  STON/O2. 3101282\n",
       "3       1  female  35.0      1      0            113803"
      ]
     },
     "execution_count": 273,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.loc[:3, 'Pclass':'Ticket']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 反向選取行列"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "透過 Python 常見的 `[::-1]` 語法，你可以輕易地改變 DataFrame 裡頭所有欄位的排列順序："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 274,
   "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>Embarked</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Parch</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Age</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Survived</th>\n",
       "      <th>PassengerId</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>S</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>22.0</td>\n",
       "      <td>male</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>C</td>\n",
       "      <td>C85</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>38.0</td>\n",
       "      <td>female</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>S</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>26.0</td>\n",
       "      <td>female</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>S</td>\n",
       "      <td>C123</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>113803</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>35.0</td>\n",
       "      <td>female</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Embarked Cabin     Fare            Ticket  Parch  SibSp   Age     Sex  \\\n",
       "0        S   NaN   7.2500         A/5 21171      0      1  22.0    male   \n",
       "1        C   C85  71.2833          PC 17599      0      1  38.0  female   \n",
       "2        S   NaN   7.9250  STON/O2. 3101282      0      0  26.0  female   \n",
       "3        S  C123  53.1000            113803      0      1  35.0  female   \n",
       "\n",
       "   Pclass  Survived  PassengerId  \n",
       "0       3         0            1  \n",
       "1       1         1            2  \n",
       "2       3         1            3  \n",
       "3       1         1            4  "
      ]
     },
     "execution_count": 274,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.loc[:3, ::-1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "同樣概念也可以運用到列（row）上面。你可以將所有樣本（samples）排序顛倒並選取其中 N 列："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 275,
   "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>Sex</th>\n",
       "      <th>Age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>890</th>\n",
       "      <td>891</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>32.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>889</th>\n",
       "      <td>890</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>26.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>888</th>\n",
       "      <td>889</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>887</th>\n",
       "      <td>888</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>19.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>886</th>\n",
       "      <td>887</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>male</td>\n",
       "      <td>27.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     PassengerId  Survived  Pclass     Sex   Age\n",
       "890          891         0       3    male  32.0\n",
       "889          890         1       1    male  26.0\n",
       "888          889         0       3  female   NaN\n",
       "887          888         1       1  female  19.0\n",
       "886          887         0       2    male  27.0"
      ]
     },
     "execution_count": 275,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.iloc[::-1, :5].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "注意我們同時使用 `:5` 來選出前 5 個欄位。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 條件選取數據"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在 pandas 裡頭最實用的選取技巧大概非遮罩（masking）莫屬了。遮罩讓 pandas 將符合特定條件的樣本回傳："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 277,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_68f75918_b060_11e9_8b6c_8c8590794fe2row0_col3 {\n",
       "            background-color:  rgb(153, 255, 51);\n",
       "        }    #T_68f75918_b060_11e9_8b6c_8c8590794fe2row0_col4 {\n",
       "            background-color:  rgb(153, 255, 51);\n",
       "        }    #T_68f75918_b060_11e9_8b6c_8c8590794fe2row1_col3 {\n",
       "            background-color:  rgb(153, 255, 51);\n",
       "        }    #T_68f75918_b060_11e9_8b6c_8c8590794fe2row1_col4 {\n",
       "            background-color:  rgb(153, 255, 51);\n",
       "        }    #T_68f75918_b060_11e9_8b6c_8c8590794fe2row2_col3 {\n",
       "            background-color:  rgb(153, 255, 51);\n",
       "        }    #T_68f75918_b060_11e9_8b6c_8c8590794fe2row2_col4 {\n",
       "            background-color:  rgb(153, 255, 51);\n",
       "        }    #T_68f75918_b060_11e9_8b6c_8c8590794fe2row3_col3 {\n",
       "            background-color:  rgb(153, 255, 51);\n",
       "        }    #T_68f75918_b060_11e9_8b6c_8c8590794fe2row3_col4 {\n",
       "            background-color:  rgb(153, 255, 51);\n",
       "        }    #T_68f75918_b060_11e9_8b6c_8c8590794fe2row4_col3 {\n",
       "            background-color:  rgb(153, 255, 51);\n",
       "        }    #T_68f75918_b060_11e9_8b6c_8c8590794fe2row4_col4 {\n",
       "            background-color:  rgb(153, 255, 51);\n",
       "        }</style><table id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2\" ><thead>    <tr>        <th class=\"blank level0\" ></th>        <th class=\"col_heading level0 col0\" >PassengerId</th>        <th class=\"col_heading level0 col1\" >Survived</th>        <th class=\"col_heading level0 col2\" >Pclass</th>        <th class=\"col_heading level0 col3\" >Sex</th>        <th class=\"col_heading level0 col4\" >Age</th>        <th class=\"col_heading level0 col5\" >SibSp</th>        <th class=\"col_heading level0 col6\" >Parch</th>        <th class=\"col_heading level0 col7\" >Ticket</th>        <th class=\"col_heading level0 col8\" >Fare</th>        <th class=\"col_heading level0 col9\" >Cabin</th>        <th class=\"col_heading level0 col10\" >Embarked</th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2level0_row0\" class=\"row_heading level0 row0\" >96</th>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row0_col0\" class=\"data row0 col0\" >97</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row0_col1\" class=\"data row0 col1\" >0</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row0_col2\" class=\"data row0 col2\" >1</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row0_col3\" class=\"data row0 col3\" >male</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row0_col4\" class=\"data row0 col4\" >71</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row0_col5\" class=\"data row0 col5\" >0</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row0_col6\" class=\"data row0 col6\" >0</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row0_col7\" class=\"data row0 col7\" >PC 17754</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row0_col8\" class=\"data row0 col8\" >34.6542</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row0_col9\" class=\"data row0 col9\" >A5</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row0_col10\" class=\"data row0 col10\" >C</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2level0_row1\" class=\"row_heading level0 row1\" >116</th>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row1_col0\" class=\"data row1 col0\" >117</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row1_col1\" class=\"data row1 col1\" >0</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row1_col2\" class=\"data row1 col2\" >3</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row1_col3\" class=\"data row1 col3\" >male</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row1_col4\" class=\"data row1 col4\" >70.5</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row1_col5\" class=\"data row1 col5\" >0</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row1_col6\" class=\"data row1 col6\" >0</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row1_col7\" class=\"data row1 col7\" >370369</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row1_col8\" class=\"data row1 col8\" >7.75</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row1_col9\" class=\"data row1 col9\" >nan</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row1_col10\" class=\"data row1 col10\" >Q</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2level0_row2\" class=\"row_heading level0 row2\" >493</th>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row2_col0\" class=\"data row2 col0\" >494</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row2_col1\" class=\"data row2 col1\" >0</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row2_col2\" class=\"data row2 col2\" >1</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row2_col3\" class=\"data row2 col3\" >male</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row2_col4\" class=\"data row2 col4\" >71</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row2_col5\" class=\"data row2 col5\" >0</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row2_col6\" class=\"data row2 col6\" >0</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row2_col7\" class=\"data row2 col7\" >PC 17609</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row2_col8\" class=\"data row2 col8\" >49.5042</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row2_col9\" class=\"data row2 col9\" >nan</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row2_col10\" class=\"data row2 col10\" >C</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2level0_row3\" class=\"row_heading level0 row3\" >630</th>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row3_col0\" class=\"data row3 col0\" >631</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row3_col1\" class=\"data row3 col1\" >1</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row3_col2\" class=\"data row3 col2\" >1</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row3_col3\" class=\"data row3 col3\" >male</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row3_col4\" class=\"data row3 col4\" >80</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row3_col5\" class=\"data row3 col5\" >0</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row3_col6\" class=\"data row3 col6\" >0</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row3_col7\" class=\"data row3 col7\" >27042</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row3_col8\" class=\"data row3 col8\" >30</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row3_col9\" class=\"data row3 col9\" >A23</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row3_col10\" class=\"data row3 col10\" >S</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2level0_row4\" class=\"row_heading level0 row4\" >851</th>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row4_col0\" class=\"data row4 col0\" >852</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row4_col1\" class=\"data row4 col1\" >0</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row4_col2\" class=\"data row4 col2\" >3</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row4_col3\" class=\"data row4 col3\" >male</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row4_col4\" class=\"data row4 col4\" >74</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row4_col5\" class=\"data row4 col5\" >0</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row4_col6\" class=\"data row4 col6\" >0</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row4_col7\" class=\"data row4 col7\" >347060</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row4_col8\" class=\"data row4 col8\" >7.775</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row4_col9\" class=\"data row4 col9\" >nan</td>\n",
       "                        <td id=\"T_68f75918_b060_11e9_8b6c_8c8590794fe2row4_col10\" class=\"data row4 col10\" >S</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1a34f720b8>"
      ]
     },
     "execution_count": 277,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "male_and_age_over_70 = (df.Sex == 'male') & (df.Age > 70)\n",
    "(df[male_and_age_over_70]\n",
    "    .style\n",
    "    .applymap(lambda x: 'background-color: rgb(153, 255, 51)', \n",
    "              subset=pd.IndexSlice[:, 'Sex':'Age']))\n",
    "# 跟 df[(df.Sex == 'male') & (df.Age > 70)] 結果相同"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`male_and_age_over_70` 是我們定義的一個遮罩，可以把同時符合兩個布林判斷式（大於 70 歲、男性）的樣本選取出來。上面註解有相同效果，但當存在多個判斷式時，有個準確說明遮罩意義的變數（上例的 `male_and_age_over_70`）會讓你的程式碼好懂一點。\n",
    "\n",
    "另外你也可以使用 `query` 函式來達到跟遮罩一樣的效果："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 278,
   "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>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>97</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>71.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17754</td>\n",
       "      <td>34.6542</td>\n",
       "      <td>A5</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116</th>\n",
       "      <td>117</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>70.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>370369</td>\n",
       "      <td>7.7500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>493</th>\n",
       "      <td>494</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>71.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17609</td>\n",
       "      <td>49.5042</td>\n",
       "      <td>NaN</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>630</th>\n",
       "      <td>631</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>80.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>27042</td>\n",
       "      <td>30.0000</td>\n",
       "      <td>A23</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>851</th>\n",
       "      <td>852</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>74.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>347060</td>\n",
       "      <td>7.7750</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     PassengerId  Survived  Pclass   Sex   Age  SibSp  Parch    Ticket  \\\n",
       "96            97         0       1  male  71.0      0      0  PC 17754   \n",
       "116          117         0       3  male  70.5      0      0    370369   \n",
       "493          494         0       1  male  71.0      0      0  PC 17609   \n",
       "630          631         1       1  male  80.0      0      0     27042   \n",
       "851          852         0       3  male  74.0      0      0    347060   \n",
       "\n",
       "        Fare Cabin Embarked  \n",
       "96   34.6542    A5        C  \n",
       "116   7.7500   NaN        Q  \n",
       "493  49.5042   NaN        C  \n",
       "630  30.0000   A23        S  \n",
       "851   7.7750   NaN        S  "
      ]
     },
     "execution_count": 278,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "age = 70\n",
    "df.query(\"Age > @age & Sex == 'male'\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在這個例子裡頭，你可以使用 `@` 來存取已經定義的 Python 變數 `age` 的值。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 選擇任一欄有空值的樣本"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "一個 DataFrame 裡常會有多個欄位（column），而每個欄位裡頭都有可能包含空值。\n",
    "\n",
    "有時候你會想把在**任一**欄位（column）出現過空值的樣本（row）全部取出："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 283,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_31d6affa_b061_11e9_8b6c_8c8590794fe2row0_col9 {\n",
       "            background-color:  red;\n",
       "        }    #T_31d6affa_b061_11e9_8b6c_8c8590794fe2row1_col9 {\n",
       "            background-color:  red;\n",
       "        }    #T_31d6affa_b061_11e9_8b6c_8c8590794fe2row2_col9 {\n",
       "            background-color:  red;\n",
       "        }    #T_31d6affa_b061_11e9_8b6c_8c8590794fe2row3_col4 {\n",
       "            background-color:  red;\n",
       "        }    #T_31d6affa_b061_11e9_8b6c_8c8590794fe2row3_col9 {\n",
       "            background-color:  red;\n",
       "        }    #T_31d6affa_b061_11e9_8b6c_8c8590794fe2row4_col9 {\n",
       "            background-color:  red;\n",
       "        }</style><table id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2\" ><thead>    <tr>        <th class=\"blank level0\" ></th>        <th class=\"col_heading level0 col0\" >PassengerId</th>        <th class=\"col_heading level0 col1\" >Survived</th>        <th class=\"col_heading level0 col2\" >Pclass</th>        <th class=\"col_heading level0 col3\" >Sex</th>        <th class=\"col_heading level0 col4\" >Age</th>        <th class=\"col_heading level0 col5\" >SibSp</th>        <th class=\"col_heading level0 col6\" >Parch</th>        <th class=\"col_heading level0 col7\" >Ticket</th>        <th class=\"col_heading level0 col8\" >Fare</th>        <th class=\"col_heading level0 col9\" >Cabin</th>        <th class=\"col_heading level0 col10\" >Embarked</th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row0_col0\" class=\"data row0 col0\" >1</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row0_col1\" class=\"data row0 col1\" >0</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row0_col2\" class=\"data row0 col2\" >3</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row0_col3\" class=\"data row0 col3\" >male</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row0_col4\" class=\"data row0 col4\" >22</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row0_col5\" class=\"data row0 col5\" >1</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row0_col6\" class=\"data row0 col6\" >0</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row0_col7\" class=\"data row0 col7\" >A/5 21171</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row0_col8\" class=\"data row0 col8\" >7.25</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row0_col9\" class=\"data row0 col9\" >nan</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row0_col10\" class=\"data row0 col10\" >S</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2level0_row1\" class=\"row_heading level0 row1\" >2</th>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row1_col0\" class=\"data row1 col0\" >3</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row1_col1\" class=\"data row1 col1\" >1</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row1_col2\" class=\"data row1 col2\" >3</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row1_col3\" class=\"data row1 col3\" >female</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row1_col4\" class=\"data row1 col4\" >26</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row1_col5\" class=\"data row1 col5\" >0</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row1_col6\" class=\"data row1 col6\" >0</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row1_col7\" class=\"data row1 col7\" >STON/O2. 3101282</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row1_col8\" class=\"data row1 col8\" >7.925</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row1_col9\" class=\"data row1 col9\" >nan</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row1_col10\" class=\"data row1 col10\" >S</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2level0_row2\" class=\"row_heading level0 row2\" >4</th>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row2_col0\" class=\"data row2 col0\" >5</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row2_col1\" class=\"data row2 col1\" >0</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row2_col2\" class=\"data row2 col2\" >3</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row2_col3\" class=\"data row2 col3\" >male</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row2_col4\" class=\"data row2 col4\" >35</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row2_col5\" class=\"data row2 col5\" >0</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row2_col6\" class=\"data row2 col6\" >0</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row2_col7\" class=\"data row2 col7\" >373450</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row2_col8\" class=\"data row2 col8\" >8.05</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row2_col9\" class=\"data row2 col9\" >nan</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row2_col10\" class=\"data row2 col10\" >S</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2level0_row3\" class=\"row_heading level0 row3\" >5</th>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row3_col0\" class=\"data row3 col0\" >6</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row3_col1\" class=\"data row3 col1\" >0</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row3_col2\" class=\"data row3 col2\" >3</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row3_col3\" class=\"data row3 col3\" >male</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row3_col4\" class=\"data row3 col4\" >nan</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row3_col5\" class=\"data row3 col5\" >0</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row3_col6\" class=\"data row3 col6\" >0</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row3_col7\" class=\"data row3 col7\" >330877</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row3_col8\" class=\"data row3 col8\" >8.4583</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row3_col9\" class=\"data row3 col9\" >nan</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row3_col10\" class=\"data row3 col10\" >Q</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2level0_row4\" class=\"row_heading level0 row4\" >7</th>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row4_col0\" class=\"data row4 col0\" >8</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row4_col1\" class=\"data row4 col1\" >0</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row4_col2\" class=\"data row4 col2\" >3</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row4_col3\" class=\"data row4 col3\" >male</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row4_col4\" class=\"data row4 col4\" >2</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row4_col5\" class=\"data row4 col5\" >3</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row4_col6\" class=\"data row4 col6\" >1</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row4_col7\" class=\"data row4 col7\" >349909</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row4_col8\" class=\"data row4 col8\" >21.075</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row4_col9\" class=\"data row4 col9\" >nan</td>\n",
       "                        <td id=\"T_31d6affa_b061_11e9_8b6c_8c8590794fe2row4_col10\" class=\"data row4 col10\" >S</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1a34f72be0>"
      ]
     },
     "execution_count": 283,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.isnull().any(axis=1)].head() \\\n",
    "    .style.highlight_null()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "這邊剛好所有樣本的 `Cabin` 欄位皆為空值。但倒數第 2 個樣本就算其 `Cabin` 欄不為空值，也會因為 `Age` 欄為空而被選出。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 選取或排除特定類型欄位"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "有時候你會想選取 DataFrame 裡特定數據類型（字串、數值、時間型態等）的欄位，這時你可以使用 `select_dtypes` 函式："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "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",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7.2500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>71.2833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.9250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>53.1000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8.0500</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived  Pclass   Age  SibSp  Parch     Fare\n",
       "0            1         0       3  22.0      1      0   7.2500\n",
       "1            2         1       1  38.0      1      0  71.2833\n",
       "2            3         1       3  26.0      0      0   7.9250\n",
       "3            4         1       1  35.0      1      0  53.1000\n",
       "4            5         0       3  35.0      0      0   8.0500"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.select_dtypes(include='number').head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "上面我們用一行程式碼就把所有數值欄位取出，儘管我們根本不知道有什麼欄位。而你當然也可以利用 `exclude` 參數來排除特定類型的欄位："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 287,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>foo1</td>\n",
       "      <td>2009-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>foo2</td>\n",
       "      <td>2009-01-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>foo3</td>\n",
       "      <td>2009-01-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>foo4</td>\n",
       "      <td>2009-01-06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>foo5</td>\n",
       "      <td>2009-01-07</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     A    B     C          D\n",
       "0  0.0  0.0  foo1 2009-01-01\n",
       "1  1.0  1.0  foo2 2009-01-02\n",
       "2  2.0  0.0  foo3 2009-01-05\n",
       "3  3.0  1.0  foo4 2009-01-06\n",
       "4  4.0  0.0  foo5 2009-01-07"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "A           float64\n",
       "B           float64\n",
       "C            object\n",
       "D    datetime64[ns]\n",
       "dtype: object"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>A</th>\n",
       "      <th>B</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     A    B\n",
       "0  0.0  0.0\n",
       "1  1.0  1.0\n",
       "2  2.0  0.0\n",
       "3  3.0  1.0\n",
       "4  4.0  0.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 建立一個有多種數據形態的 DataFrame\n",
    "df_mix = pd.util.testing.makeMixedDataFrame()\n",
    "display(df_mix)\n",
    "display(df_mix.dtypes)\n",
    "display(df_mix.select_dtypes(exclude=['datetime64', 'object']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "pandas 裡的函式使用上都很直覺，你可以丟入 1 個包含多個元素的 Python `list` 或是單一 `str` 作為參數輸入。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 選取所有出現在 list 內的樣本"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "很多時候針對某一個特定欄位，你會想要取出所有出現在一個 `list` 的樣本。這時候你可以使用 `isin` 函式來做到這件事情："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "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>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>44</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>female</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>SC/Paris 2123</td>\n",
       "      <td>41.5792</td>\n",
       "      <td>NaN</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>608</th>\n",
       "      <td>609</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>female</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>SC/Paris 2123</td>\n",
       "      <td>41.5792</td>\n",
       "      <td>NaN</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>685</th>\n",
       "      <td>686</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>male</td>\n",
       "      <td>25.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>SC/Paris 2123</td>\n",
       "      <td>41.5792</td>\n",
       "      <td>NaN</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>701</th>\n",
       "      <td>702</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17475</td>\n",
       "      <td>26.2875</td>\n",
       "      <td>E24</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     PassengerId  Survived  Pclass     Sex   Age  SibSp  Parch         Ticket  \\\n",
       "43            44         1       2  female   3.0      1      2  SC/Paris 2123   \n",
       "608          609         1       2  female  22.0      1      2  SC/Paris 2123   \n",
       "685          686         0       2    male  25.0      1      2  SC/Paris 2123   \n",
       "701          702         1       1    male  35.0      0      0       PC 17475   \n",
       "\n",
       "        Fare Cabin Embarked  \n",
       "43   41.5792   NaN        C  \n",
       "608  41.5792   NaN        C  \n",
       "685  41.5792   NaN        C  \n",
       "701  26.2875   E24        S  "
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tickets = [\"SC/Paris 2123\", \"PC 17475\"]\n",
    "df[df.Ticket.isin(tickets)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 選取某欄位為 top-k 值的樣本"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "很多時候你會想選取在某個欄位中前 k 大的所有樣本。這時你可以先利用 `value_counts` 函式找出該欄位前 k 多的值："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['347082', '1601', 'CA. 2343'], dtype='object')"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_k = 3\n",
    "top_tickets = df.Ticket.value_counts()[:top_k]\n",
    "top_tickets.index"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "這邊我們以欄位 `Ticket` 為例。另外你也可以使用 [pandas.Series](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.Series.html) 裡的 `nlargest` 函式取得相同結果："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['347082', '1601', 'CA. 2343'], dtype='object')"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.Ticket.value_counts().nlargest(top_k).index"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "接著利用上小節看過的 `isin` 函式就能輕鬆取得 `Ticket`欄位值為前 k 大值的樣本："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "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>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>14</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>39.0</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>347082</td>\n",
       "      <td>31.2750</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74</th>\n",
       "      <td>75</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>32.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1601</td>\n",
       "      <td>56.4958</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>120</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>347082</td>\n",
       "      <td>31.2750</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>159</th>\n",
       "      <td>160</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>CA. 2343</td>\n",
       "      <td>69.5500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>170</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>28.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1601</td>\n",
       "      <td>56.4958</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     PassengerId  Survived  Pclass     Sex   Age  SibSp  Parch    Ticket  \\\n",
       "13            14         0       3    male  39.0      1      5    347082   \n",
       "74            75         1       3    male  32.0      0      0      1601   \n",
       "119          120         0       3  female   2.0      4      2    347082   \n",
       "159          160         0       3    male   NaN      8      2  CA. 2343   \n",
       "169          170         0       3    male  28.0      0      0      1601   \n",
       "\n",
       "        Fare Cabin Embarked  \n",
       "13   31.2750   NaN        S  \n",
       "74   56.4958   NaN        S  \n",
       "119  31.2750   NaN        S  \n",
       "159  69.5500   NaN        S  \n",
       "169  56.4958   NaN        S  "
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.Ticket.isin(top_tickets.index)].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 找出符合特定字串的樣本"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "有時你會想要對一個字串欄位做正規表示式（regular expression），取出符合某個 pattern 的所有樣本。\n",
    "\n",
    "這時你可以使用 `str` 底下的 `contains` 函式："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Braund, Mr. Owen Harris</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Allen, Mr. William Henry</td>\n",
       "      <td>male</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>373450</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Moran, Mr. James</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>330877</td>\n",
       "      <td>8.4583</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>McCarthy, Mr. Timothy J</td>\n",
       "      <td>male</td>\n",
       "      <td>54.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>17463</td>\n",
       "      <td>51.8625</td>\n",
       "      <td>E46</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>13</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Saundercock, Mr. William Henry</td>\n",
       "      <td>male</td>\n",
       "      <td>20.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5. 2151</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                            Name   Sex   Age  \\\n",
       "0             1         0       3         Braund, Mr. Owen Harris  male  22.0   \n",
       "4             5         0       3        Allen, Mr. William Henry  male  35.0   \n",
       "5             6         0       3                Moran, Mr. James  male   NaN   \n",
       "6             7         0       1         McCarthy, Mr. Timothy J  male  54.0   \n",
       "12           13         0       3  Saundercock, Mr. William Henry  male  20.0   \n",
       "\n",
       "    SibSp  Parch     Ticket     Fare Cabin Embarked  \n",
       "0       1      0  A/5 21171   7.2500   NaN        S  \n",
       "4       0      0     373450   8.0500   NaN        S  \n",
       "5       0      0     330877   8.4583   NaN        Q  \n",
       "6       0      0      17463  51.8625   E46        S  \n",
       "12      0      0  A/5. 2151   8.0500   NaN        S  "
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('http://bit.ly/kaggletrain')\n",
    "df[df.Name.str.contains(\"Mr\\.\")].head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "這邊我們將所有 `Name` 欄位值裡包含 `Mr.` 的樣本取出。注意 `contains` 函式接受的是正規表示式，因此需要將 `.` 轉換成 `\\.`。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 使用正規表示式選取數據"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "有時候你會想要依照一些規則來選取 DataFrame 裡頭的值、索引或是欄位，尤其是在處理跟時間序列相關的數據："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2000-01-01</th>\n",
       "      <td>-0.143945</td>\n",
       "      <td>-0.020783</td>\n",
       "      <td>1.434651</td>\n",
       "      <td>1.044208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-01-08</th>\n",
       "      <td>-0.901508</td>\n",
       "      <td>-1.082932</td>\n",
       "      <td>-1.035743</td>\n",
       "      <td>0.334431</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-01-15</th>\n",
       "      <td>-1.544868</td>\n",
       "      <td>1.183136</td>\n",
       "      <td>-0.913117</td>\n",
       "      <td>-1.199199</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-01-22</th>\n",
       "      <td>0.604091</td>\n",
       "      <td>-0.233628</td>\n",
       "      <td>-0.348109</td>\n",
       "      <td>1.082138</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-01-29</th>\n",
       "      <td>0.062112</td>\n",
       "      <td>1.565870</td>\n",
       "      <td>-0.791369</td>\n",
       "      <td>1.017766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-02-05</th>\n",
       "      <td>-0.013661</td>\n",
       "      <td>1.015528</td>\n",
       "      <td>-0.420123</td>\n",
       "      <td>-0.513559</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-02-12</th>\n",
       "      <td>0.404094</td>\n",
       "      <td>0.486909</td>\n",
       "      <td>-0.815937</td>\n",
       "      <td>0.743381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-02-19</th>\n",
       "      <td>-0.288886</td>\n",
       "      <td>2.560776</td>\n",
       "      <td>-0.864528</td>\n",
       "      <td>0.727740</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-02-26</th>\n",
       "      <td>0.975203</td>\n",
       "      <td>-0.551452</td>\n",
       "      <td>0.531635</td>\n",
       "      <td>-0.595716</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-03-04</th>\n",
       "      <td>-0.045714</td>\n",
       "      <td>0.137413</td>\n",
       "      <td>2.187056</td>\n",
       "      <td>1.164371</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "2000-01-01 -0.143945 -0.020783  1.434651  1.044208\n",
       "2000-01-08 -0.901508 -1.082932 -1.035743  0.334431\n",
       "2000-01-15 -1.544868  1.183136 -0.913117 -1.199199\n",
       "2000-01-22  0.604091 -0.233628 -0.348109  1.082138\n",
       "2000-01-29  0.062112  1.565870 -0.791369  1.017766\n",
       "2000-02-05 -0.013661  1.015528 -0.420123 -0.513559\n",
       "2000-02-12  0.404094  0.486909 -0.815937  0.743381\n",
       "2000-02-19 -0.288886  2.560776 -0.864528  0.727740\n",
       "2000-02-26  0.975203 -0.551452  0.531635 -0.595716\n",
       "2000-03-04 -0.045714  0.137413  2.187056  1.164371"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_date = pd.util.testing.makeTimeDataFrame(freq='7D')\n",
    "df_date.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "假設你想將所有索引在 2000 年 2 月內的樣本取出，則可以透過 `filter` 函式達成這個目的："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2000-02-05</th>\n",
       "      <td>-0.013661</td>\n",
       "      <td>1.015528</td>\n",
       "      <td>-0.420123</td>\n",
       "      <td>-0.513559</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-02-12</th>\n",
       "      <td>0.404094</td>\n",
       "      <td>0.486909</td>\n",
       "      <td>-0.815937</td>\n",
       "      <td>0.743381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-02-19</th>\n",
       "      <td>-0.288886</td>\n",
       "      <td>2.560776</td>\n",
       "      <td>-0.864528</td>\n",
       "      <td>0.727740</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-02-26</th>\n",
       "      <td>0.975203</td>\n",
       "      <td>-0.551452</td>\n",
       "      <td>0.531635</td>\n",
       "      <td>-0.595716</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "2000-02-05 -0.013661  1.015528 -0.420123 -0.513559\n",
       "2000-02-12  0.404094  0.486909 -0.815937  0.743381\n",
       "2000-02-19 -0.288886  2.560776 -0.864528  0.727740\n",
       "2000-02-26  0.975203 -0.551452  0.531635 -0.595716"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_date.filter(regex=\"2000-02.*\", axis=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`filter` 函式本身功能十分強大，有興趣的讀者可以[閱讀官方文件進一步了解其用法](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.filter.html)。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 選取從某時間點開始的區間樣本"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在處理時間數據時，很多時候你會想要針對某個起始時間挑出前 t 個時間點的樣本。\n",
    "\n",
    "讓我們再看一次剛剛建立的 DataFrame："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2000-01-01</th>\n",
       "      <td>-0.143945</td>\n",
       "      <td>-0.020783</td>\n",
       "      <td>1.434651</td>\n",
       "      <td>1.044208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-01-08</th>\n",
       "      <td>-0.901508</td>\n",
       "      <td>-1.082932</td>\n",
       "      <td>-1.035743</td>\n",
       "      <td>0.334431</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-01-15</th>\n",
       "      <td>-1.544868</td>\n",
       "      <td>1.183136</td>\n",
       "      <td>-0.913117</td>\n",
       "      <td>-1.199199</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-01-22</th>\n",
       "      <td>0.604091</td>\n",
       "      <td>-0.233628</td>\n",
       "      <td>-0.348109</td>\n",
       "      <td>1.082138</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-01-29</th>\n",
       "      <td>0.062112</td>\n",
       "      <td>1.565870</td>\n",
       "      <td>-0.791369</td>\n",
       "      <td>1.017766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-02-05</th>\n",
       "      <td>-0.013661</td>\n",
       "      <td>1.015528</td>\n",
       "      <td>-0.420123</td>\n",
       "      <td>-0.513559</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-02-12</th>\n",
       "      <td>0.404094</td>\n",
       "      <td>0.486909</td>\n",
       "      <td>-0.815937</td>\n",
       "      <td>0.743381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-02-19</th>\n",
       "      <td>-0.288886</td>\n",
       "      <td>2.560776</td>\n",
       "      <td>-0.864528</td>\n",
       "      <td>0.727740</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "2000-01-01 -0.143945 -0.020783  1.434651  1.044208\n",
       "2000-01-08 -0.901508 -1.082932 -1.035743  0.334431\n",
       "2000-01-15 -1.544868  1.183136 -0.913117 -1.199199\n",
       "2000-01-22  0.604091 -0.233628 -0.348109  1.082138\n",
       "2000-01-29  0.062112  1.565870 -0.791369  1.017766\n",
       "2000-02-05 -0.013661  1.015528 -0.420123 -0.513559\n",
       "2000-02-12  0.404094  0.486909 -0.815937  0.743381\n",
       "2000-02-19 -0.288886  2.560776 -0.864528  0.727740"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_date.head(8)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在索引為時間型態的情況下，如果你想要把前 3 週的樣本取出，可以使用 `first` 函式："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2000-01-01</th>\n",
       "      <td>-0.143945</td>\n",
       "      <td>-0.020783</td>\n",
       "      <td>1.434651</td>\n",
       "      <td>1.044208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-01-08</th>\n",
       "      <td>-0.901508</td>\n",
       "      <td>-1.082932</td>\n",
       "      <td>-1.035743</td>\n",
       "      <td>0.334431</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-01-15</th>\n",
       "      <td>-1.544868</td>\n",
       "      <td>1.183136</td>\n",
       "      <td>-0.913117</td>\n",
       "      <td>-1.199199</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   A         B         C         D\n",
       "2000-01-01 -0.143945 -0.020783  1.434651  1.044208\n",
       "2000-01-08 -0.901508 -1.082932 -1.035743  0.334431\n",
       "2000-01-15 -1.544868  1.183136 -0.913117 -1.199199"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_date.first('3W')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "十分方便的函式。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## 基本數據處理與轉換"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在了解如何選取想要的數據以後，你可以透過這節的介紹來熟悉 pandas 裡一些常見的數據處理方式。這章節也是我認為使用 pandas 處理數據時最令人愉快的部分之一。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 對某一軸套用相同運算"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你時常會需要對 DataFrame 裡頭的每一個欄位（縱軸）或是每一列（橫軸）做相同的運算。\n",
    "\n",
    "比方說你想將鐵達尼號資料集內的 `Survived` 數值欄位轉換成人類容易理解的字串："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "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>存活</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Sex</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>死亡</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>倖存</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>倖存</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>倖存</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>死亡</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>死亡</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  存活  Survived  Pclass     Sex\n",
       "0            1  死亡         0       3    male\n",
       "1            2  倖存         1       1  female\n",
       "2            3  倖存         1       3  female\n",
       "3            4  倖存         1       1  female\n",
       "4            5  死亡         0       3    male\n",
       "5            6  死亡         0       3    male"
      ]
     },
     "execution_count": 194,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 重新讀取鐵達尼號數據\n",
    "df_titanic = pd.read_csv('http://bit.ly/kaggletrain')\n",
    "df_titanic = df_titanic.drop(\"Name\", axis=1)\n",
    "\n",
    "# 複製一份副本 DataFrame\n",
    "df = df_titanic.copy()\n",
    "columns = df.columns.tolist()[:4]\n",
    "\n",
    "# 好戲登場\n",
    "new_col = '存活'\n",
    "columns.insert(1, new_col)  # 調整欄位順序用\n",
    "df[new_col] = df.Survived.apply(lambda x: '倖存' if x else '死亡')\n",
    "df.loc[:5, columns]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "透過 `apply` 函式，我們把一個匿名函式 `lambda` 套用到整個 `df.Survived` Series 之上，並以此建立一個新的 `存活` 欄位。\n",
    "\n",
    "值得一提的是當你需要追加新的欄位但又不想影響到原始 DataFrame 時，可以使用 `copy` 函式複製一份副本另行操作。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 對每一個樣本做自定義運算"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "上小節我們用 `apply` 函式對 DataFrame 裡頭的**某個 Series **做運算並生成新欄位：\n",
    "\n",
    "```python\n",
    "df[new_col] = df.Survived.apply(...\n",
    "```\n",
    "\n",
    "不過你時常會想要把樣本（row）裡頭的**多個**欄位一次取出做運算並產生一個新的值。這時你可以自定義一個 Python function 並將 `apply` 函式套用到整個 DataFrame 之上："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "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>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>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     Sex   Age  SibSp  Parch  \\\n",
       "0            1         0       3    male  22.0      1      0   \n",
       "1            2         1       1  female  38.0      1      0   \n",
       "2            3         1       3  female  26.0      0      0   \n",
       "3            4         1       1  female  35.0      1      0   \n",
       "4            5         0       3    male  35.0      0      0   \n",
       "\n",
       "             Ticket     Fare Cabin Embarked  \n",
       "0         A/5 21171   7.2500   NaN        S  \n",
       "1          PC 17599  71.2833   C85        C  \n",
       "2  STON/O2. 3101282   7.9250   NaN        S  \n",
       "3            113803  53.1000  C123        S  \n",
       "4            373450   8.0500   NaN        S  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "      <th>描述</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>一名 22.0 歲的男性</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "      <td>一名 38.0 歲的女性</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>一名 26.0 歲的女性</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "      <td>一名 35.0 歲的女性</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>male</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>373450</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>一名 35.0 歲的男性</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Sex   Age  SibSp  Parch            Ticket     Fare Cabin Embarked  \\\n",
       "0    male  22.0      1      0         A/5 21171   7.2500   NaN        S   \n",
       "1  female  38.0      1      0          PC 17599  71.2833   C85        C   \n",
       "2  female  26.0      0      0  STON/O2. 3101282   7.9250   NaN        S   \n",
       "3  female  35.0      1      0            113803  53.1000  C123        S   \n",
       "4    male  35.0      0      0            373450   8.0500   NaN        S   \n",
       "\n",
       "             描述  \n",
       "0  一名 22.0 歲的男性  \n",
       "1  一名 38.0 歲的女性  \n",
       "2  一名 26.0 歲的女性  \n",
       "3  一名 35.0 歲的女性  \n",
       "4  一名 35.0 歲的男性  "
      ]
     },
     "execution_count": 207,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df_titanic.copy()\n",
    "display(df.head())\n",
    "\n",
    "# apply custom function 可以說是 pandas 裡最重要的技巧之一\n",
    "d = {'male': '男性', 'female': '女性'} \n",
    "def generate_desc(row):\n",
    "    return f\"一名 {row['Age']} 歲的{d[row['Sex']]}\"\n",
    "\n",
    "df['描述'] = df.apply(generate_desc, axis=1)\n",
    "df.loc[:4, 'Sex':]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "此例中 `apply` 函式將 `generate_desc` 函式**個別**套用到 DataFrame 裡頭的每一個樣本（row），結合 `Sex` 及 `Age` 兩欄位資訊，生成新的 `描述`。\n",
    "\n",
    "當然，將 `axis` 設置為 `0` 則可以對每一個欄位分別套用自定義的 Python function。 "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 將連續數值轉換成分類數據"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "有時你會想把一個連續數值（numerical）的欄位分成多個 groups 以方便對每個 groups 做統計。這時候你可以使用 `pd.cut` 函式："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 290,
   "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>Sex</th>\n",
       "      <th>年齡區間</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>631</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>族群 10</td>\n",
       "      <td>80.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>27042</td>\n",
       "      <td>30.0000</td>\n",
       "      <td>A23</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>852</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>族群 10</td>\n",
       "      <td>74.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>347060</td>\n",
       "      <td>7.7750</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>55</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>族群 9</td>\n",
       "      <td>65.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>113509</td>\n",
       "      <td>61.9792</td>\n",
       "      <td>B30</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>97</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>族群 9</td>\n",
       "      <td>71.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17754</td>\n",
       "      <td>34.6542</td>\n",
       "      <td>A5</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>494</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>族群 9</td>\n",
       "      <td>71.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17609</td>\n",
       "      <td>49.5042</td>\n",
       "      <td>NaN</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>117</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>族群 9</td>\n",
       "      <td>70.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>370369</td>\n",
       "      <td>7.7500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived  Pclass   Sex   年齡區間   Age  SibSp  Parch    Ticket  \\\n",
       "0          631         1       1  male  族群 10  80.0      0      0     27042   \n",
       "1          852         0       3  male  族群 10  74.0      0      0    347060   \n",
       "2           55         0       1  male   族群 9  65.0      0      1    113509   \n",
       "3           97         0       1  male   族群 9  71.0      0      0  PC 17754   \n",
       "4          494         0       1  male   族群 9  71.0      0      0  PC 17609   \n",
       "5          117         0       3  male   族群 9  70.5      0      0    370369   \n",
       "\n",
       "      Fare Cabin Embarked  \n",
       "0  30.0000   A23        S  \n",
       "1   7.7750   NaN        S  \n",
       "2  61.9792   B30        C  \n",
       "3  34.6542    A5        C  \n",
       "4  49.5042   NaN        C  \n",
       "5   7.7500   NaN        Q  "
      ]
     },
     "execution_count": 290,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df_titanic.copy()\n",
    "\n",
    "# 為了方便比較新舊欄位\n",
    "columns = df.columns.tolist()\n",
    "new_col = '年齡區間'\n",
    "columns.insert(4, new_col)\n",
    "\n",
    "# 將 numerical 轉換成 categorical 欄位\n",
    "labels = [f'族群 {i}' for i in range(1, 11)]\n",
    "df[new_col] = pd.cut(x=df.Age, \n",
    "                     bins=10, \n",
    "                     labels=labels)\n",
    "\n",
    "# 可以排序切割後的 categorical 欄位\n",
    "(df.sort_values(new_col, ascending=False)\n",
    "   .reset_index()\n",
    "   .loc[:5, columns]\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "如上所示，使用 `pd.cut` 函式建立出來的每個分類 `族群 X` 有大小之分，因此你可以輕易地使用 `sort_values` 函式排序樣本。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CategoricalDtype(categories=['族群 1', '族群 2', '族群 3', '族群 4', '族群 5', '族群 6', '族群 7',\n",
       "                  '族群 8', '族群 9', '族群 10'],\n",
       "                 ordered=True)"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[new_col].dtype"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 將 DataFrame 隨機切成兩個子集"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "有時你會想將手上的 DataFrame 隨機切成兩個獨立的子集。選取其中一個子集來訓練機器學習模型是一個常見的情境。\n",
    "\n",
    "要做到這件事情有很多種方法，你可以使用 [scikit-learn 的 `train_test_split`](https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.train_test_split.html) 或是 numpy 的 `np.random.randn`，但假如你想要純 pandas 解法，可以使用 `sample` 函式："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "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>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>294</th>\n",
       "      <td>295</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>24.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>349233</td>\n",
       "      <td>7.8958</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199</th>\n",
       "      <td>200</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>female</td>\n",
       "      <td>24.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>248747</td>\n",
       "      <td>13.0000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>864</th>\n",
       "      <td>865</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>male</td>\n",
       "      <td>24.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>233866</td>\n",
       "      <td>13.0000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>625</th>\n",
       "      <td>626</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>male</td>\n",
       "      <td>61.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>36963</td>\n",
       "      <td>32.3208</td>\n",
       "      <td>D50</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>638</th>\n",
       "      <td>639</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>41.0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>3101295</td>\n",
       "      <td>39.6875</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     PassengerId  Survived  Pclass     Sex   Age  SibSp  Parch   Ticket  \\\n",
       "294          295         0       3    male  24.0      0      0   349233   \n",
       "199          200         0       2  female  24.0      0      0   248747   \n",
       "864          865         0       2    male  24.0      0      0   233866   \n",
       "625          626         0       1    male  61.0      0      0    36963   \n",
       "638          639         0       3  female  41.0      0      5  3101295   \n",
       "\n",
       "        Fare Cabin Embarked  \n",
       "294   7.8958   NaN        S  \n",
       "199  13.0000   NaN        S  \n",
       "864  13.0000   NaN        S  \n",
       "625  32.3208   D50        S  \n",
       "638  39.6875   NaN        S  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>PP 9549</td>\n",
       "      <td>16.7000</td>\n",
       "      <td>G6</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>58.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>113783</td>\n",
       "      <td>26.5500</td>\n",
       "      <td>C103</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>13</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>20.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5. 2151</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>15</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>350406</td>\n",
       "      <td>7.8542</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>17</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>382652</td>\n",
       "      <td>29.1250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    PassengerId  Survived  Pclass     Sex   Age  SibSp  Parch     Ticket  \\\n",
       "10           11         1       3  female   4.0      1      1    PP 9549   \n",
       "11           12         1       1  female  58.0      0      0     113783   \n",
       "12           13         0       3    male  20.0      0      0  A/5. 2151   \n",
       "14           15         0       3  female  14.0      0      0     350406   \n",
       "16           17         0       3    male   2.0      4      1     382652   \n",
       "\n",
       "       Fare Cabin Embarked  \n",
       "10  16.7000    G6        S  \n",
       "11  26.5500  C103        S  \n",
       "12   8.0500   NaN        S  \n",
       "14   7.8542   NaN        S  \n",
       "16  29.1250   NaN        Q  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "各 DataFrame 大小： 891 713 178\n"
     ]
    }
   ],
   "source": [
    "df_train = df_titanic.sample(frac=0.8, random_state=5566)\n",
    "df_test = df_titanic.drop(df_train.index)\n",
    "\n",
    "# 顯示結果，無特殊操作\n",
    "display(df_train.head())\n",
    "display(df_test.head())\n",
    "print('各 DataFrame 大小：', \n",
    "      len(df_titanic), len(df_train), len(df_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "這個解法的前提是原來的 DataFrame `df_titanic` 裡頭的索引是獨一無二的。另外記得設定 `random_state` 以方便別人重現你的結果。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 用 SQL 的方式合併兩個 DataFrames"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "很多時候你會想要將兩個 DataFrames 依照某個共通的欄位（鍵值）合併成單一 DataFrame 以整合資訊。\n",
    "\n",
    "比方說給定以下兩個 DataFrames："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "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>state</th>\n",
       "      <th>city</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>密蘇里州</td>\n",
       "      <td>堪薩斯城</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>亞利桑那州</td>\n",
       "      <td>鳳凰城</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>肯塔基州</td>\n",
       "      <td>路易維爾</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>紐約州</td>\n",
       "      <td>紐約市</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   state  city\n",
       "0   密蘇里州  堪薩斯城\n",
       "1  亞利桑那州   鳳凰城\n",
       "2   肯塔基州  路易維爾\n",
       "3    紐約州   紐約市"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>city</th>\n",
       "      <th>population</th>\n",
       "      <th>feature</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>路易維爾</td>\n",
       "      <td>741096</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>堪薩斯城</td>\n",
       "      <td>481420</td>\n",
       "      <td>b</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>鳳凰城</td>\n",
       "      <td>4039182</td>\n",
       "      <td>c</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   city  population feature\n",
       "0  路易維爾      741096       a\n",
       "1  堪薩斯城      481420       b\n",
       "2   鳳凰城     4039182       c"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_city = pd.DataFrame({\n",
    "    'state': ['密蘇里州', '亞利桑那州', '肯塔基州', '紐約州'],\n",
    "    'city': ['堪薩斯城', '鳳凰城', '路易維爾', '紐約市']\n",
    "})\n",
    "df_info = pd.DataFrame({\n",
    "    'city': ['路易維爾', '堪薩斯城', '鳳凰城'],\n",
    "    'population': [741096, 481420, 4039182],\n",
    "    'feature': list('abc')})\n",
    "\n",
    "display(df_city)\n",
    "display(df_info)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "DataFrame `df_city` 包含了幾個美國城市以及其對應的州名（state）；DataFrame `df_info` 則包含城市名稱以及一些數據。如果你想將這兩個 DataFrames 合併（merge），可以使用非常方便的 `merge` 函式："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "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>state</th>\n",
       "      <th>city</th>\n",
       "      <th>population</th>\n",
       "      <th>feature</th>\n",
       "      <th>_merge</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>密蘇里州</td>\n",
       "      <td>堪薩斯城</td>\n",
       "      <td>481420.0</td>\n",
       "      <td>b</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>亞利桑那州</td>\n",
       "      <td>鳳凰城</td>\n",
       "      <td>4039182.0</td>\n",
       "      <td>c</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>肯塔基州</td>\n",
       "      <td>路易維爾</td>\n",
       "      <td>741096.0</td>\n",
       "      <td>a</td>\n",
       "      <td>both</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>紐約州</td>\n",
       "      <td>紐約市</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>left_only</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   state  city  population feature     _merge\n",
       "0   密蘇里州  堪薩斯城    481420.0       b       both\n",
       "1  亞利桑那州   鳳凰城   4039182.0       c       both\n",
       "2   肯塔基州  路易維爾    741096.0       a       both\n",
       "3    紐約州   紐約市         NaN     NaN  left_only"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.merge(left=df_city,\n",
    "         right=df_info,\n",
    "         how=\"left\", # left outer join\n",
    "         on=\"city\", # 透過此欄位合併\n",
    "         indicator=True # 顯示結果中每一列的來源\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "沒錯，`merge` 函式運作方式就像 [SQL](https://leemeng.tw/why-you-need-to-learn-sql-as-a-data-scientist.html) 一樣，可以讓你透過更改 `how` 參數來做：\n",
    "- `left`：left outer join\n",
    "- `right`：right outer join\n",
    "- `outer`: full outer join\n",
    "- `inner`： inner join\n",
    "\n",
    "注意合併後的 DataFrame 的最後一列：因為是 left join，就算右側的 `df_info` 裡頭並沒有紐約市的資訊，我們也能把該城市保留在 `merge` 後的結果。你還可以透過 `indicator=True` 的方式讓 pandas 幫我們新增一個 `_merge` 欄位，輕鬆了解紐約市只存在左側的 `df_city` 裡。\n",
    "\n",
    "`merge` 函式強大之處在於能跟 SQL 一樣為我們抽象化如何合併兩個 DataFrames 的運算。如果你想要了解 SQL 跟 Python 本質上的差異，可以參考[為何資料科學家需要學習 SQL](https://leemeng.tw/why-you-need-to-learn-sql-as-a-data-scientist.html) 一文。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 存取並操作每一個樣本"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "我們前面看過，雖然一般可以直接使用 [`apply` 函式](#對某一軸套用相同運算)來對每個樣本作運算，有時候你就是會想用 for 迴圈的方式把每個樣本取出處理。\n",
    "\n",
    "這種時候你可以用 `itertuples` 函式："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "堪薩斯城是密蘇里州裡頭的一個城市\n",
      "鳳凰城是亞利桑那州裡頭的一個城市\n",
      "路易維爾是肯塔基州裡頭的一個城市\n",
      "紐約市是紐約州裡頭的一個城市\n"
     ]
    }
   ],
   "source": [
    "for row in df_city.itertuples(name='City'):\n",
    "    print(f'{row.city}是{row.state}裡頭的一個城市')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "顧名思義，`itertuples` 函式回傳的是 Python [namedtuple](https://docs.python.org/3/library/collections.html#collections.namedtuple)，也是一個你應該已經很熟悉的資料型態："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from collections import namedtuple\n",
    "\n",
    "City = namedtuple('City', ['Index', 'state', 'city'])\n",
    "c = City(3, '紐約州', '紐約市')\n",
    "c == row"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "toc-hr-collapsed": true,
    "toc-nb-collapsed": true
   },
   "source": [
    "## 簡單匯總 & 分析數據"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在對數據做些基本處理以後，你會想要從手上的 DataFrame 匯總或整理出一些有用的統計數據。本節介紹一些常用的數據匯總技巧。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 取出某欄位 top k 的值"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "這你在[選取某欄位為 top-k 值的樣本](#選取某欄位為-top-k-值的樣本)小節應該就看過了。\n",
    "\n",
    "但因為這個使用情境實在太常出現，讓我們再次嘗試將鐵達尼號數據集裡頭 `Ticket` 欄位最常出現的值取出："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 292,
   "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>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>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     Sex   Age  SibSp  Parch  \\\n",
       "0            1         0       3    male  22.0      1      0   \n",
       "1            2         1       1  female  38.0      1      0   \n",
       "2            3         1       3  female  26.0      0      0   \n",
       "3            4         1       1  female  35.0      1      0   \n",
       "4            5         0       3    male  35.0      0      0   \n",
       "\n",
       "             Ticket     Fare Cabin Embarked  \n",
       "0         A/5 21171   7.2500   NaN        S  \n",
       "1          PC 17599  71.2833   C85        C  \n",
       "2  STON/O2. 3101282   7.9250   NaN        S  \n",
       "3            113803  53.1000  C123        S  \n",
       "4            373450   8.0500   NaN        S  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>Ticket</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>347082</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1601</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>CA. 2343</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3101295</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>CA 2144</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      index  Ticket\n",
       "0    347082       7\n",
       "1      1601       7\n",
       "2  CA. 2343       7\n",
       "3   3101295       6\n",
       "4   CA 2144       6"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df = df_titanic.copy()\n",
    "display(df.head())\n",
    "display(df.Ticket.value_counts().head(5).reset_index())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`value_counts` 函式預設就會把欄位裡頭的值依照出現頻率**由高到低**排序，因此搭配 `head` 函式就可以把最常出現的 top k 值選出。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 一行描述數值欄位"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "當你想要快速了解 DataFrame 裡所有數值欄位的統計數據（最小值、最大值、平均和中位數等）時可以使用 `describe` 函式："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "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",
       "      <th>count</th>\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",
       "      <th>mean</th>\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",
       "      <th>std</th>\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",
       "      <th>min</th>\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",
       "      <th>25%</th>\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",
       "      <th>50%</th>\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",
       "      <th>75%</th>\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",
       "      <th>max</th>\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": 175,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你也可以用[取得想要關注的數據](#取得想要關注的數據_1)一節的技巧來選取自己關心的統計數據："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "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>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.383838</td>\n",
       "      <td>2.308642</td>\n",
       "      <td>29.699118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.486592</td>\n",
       "      <td>0.836071</td>\n",
       "      <td>14.526497</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Survived    Pclass        Age\n",
       "mean  0.383838  2.308642  29.699118\n",
       "std   0.486592  0.836071  14.526497"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe().loc[['mean', 'std'], 'Survived':'Age']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 找出欄位裡所有出現過的值"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "針對特定欄位使用 `unique` 函式即可："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['male', 'female'], dtype=object)"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.Sex.unique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 分組匯總結果"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "很多時候你會想要把 DataFrame 裡頭的樣本依照某些特性分門別類，並依此匯總各組（group）的統計數據。這種時候你可以用 `groupby` 函式。\n",
    "\n",
    "讓我們再次拿出鐵達尼號數據集："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "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>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>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     Sex   Age  SibSp  Parch  \\\n",
       "0            1         0       3    male  22.0      1      0   \n",
       "1            2         1       1  female  38.0      1      0   \n",
       "2            3         1       3  female  26.0      0      0   \n",
       "3            4         1       1  female  35.0      1      0   \n",
       "4            5         0       3    male  35.0      0      0   \n",
       "\n",
       "             Ticket     Fare Cabin Embarked  \n",
       "0         A/5 21171   7.2500   NaN        S  \n",
       "1          PC 17599  71.2833   C85        C  \n",
       "2  STON/O2. 3101282   7.9250   NaN        S  \n",
       "3            113803  53.1000  C123        S  \n",
       "4            373450   8.0500   NaN        S  "
      ]
     },
     "execution_count": 208,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df_titanic.copy()\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你可以將所有乘客（列）依照它們的 `Pclass` 欄位值分組，並計算每組裡頭乘客們的平均年齡："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pclass\n",
       "1    38.233441\n",
       "2    29.877630\n",
       "3    25.140620\n",
       "Name: Age, dtype: float64"
      ]
     },
     "execution_count": 177,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby(\"Pclass\").Age.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你也可以搭配剛剛看過的 `describe` 函式來匯總各組的統計數據："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Sex</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>female</th>\n",
       "      <td>314.0</td>\n",
       "      <td>0.742038</td>\n",
       "      <td>0.438211</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>male</th>\n",
       "      <td>577.0</td>\n",
       "      <td>0.188908</td>\n",
       "      <td>0.391775</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        count      mean       std  min  25%  50%  75%  max\n",
       "Sex                                                       \n",
       "female  314.0  0.742038  0.438211  0.0  0.0  1.0  1.0  1.0\n",
       "male    577.0  0.188908  0.391775  0.0  0.0  0.0  0.0  1.0"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby(\"Sex\").Survived.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你也可以依照多個欄位分組，並利用 `size` 函式迅速地取得各組包含的樣本數："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "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>Pclass</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Sex</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>female</th>\n",
       "      <td>94</td>\n",
       "      <td>76</td>\n",
       "      <td>144</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>male</th>\n",
       "      <td>122</td>\n",
       "      <td>108</td>\n",
       "      <td>347</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Pclass    1    2    3\n",
       "Sex                  \n",
       "female   94   76  144\n",
       "male    122  108  347"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby([\"Sex\", 'Pclass']).size().unstack()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你也可以用 `agg` 函式（**agg**regate，匯總）搭配 `groupby` 函式來將每一組樣本依照多種方式匯總："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "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></th>\n",
       "      <th>min</th>\n",
       "      <th>max</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Sex</th>\n",
       "      <th>Pclass</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">female</th>\n",
       "      <th>1</th>\n",
       "      <td>2.00</td>\n",
       "      <td>63.0</td>\n",
       "      <td>85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.00</td>\n",
       "      <td>57.0</td>\n",
       "      <td>74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.75</td>\n",
       "      <td>63.0</td>\n",
       "      <td>102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">male</th>\n",
       "      <th>1</th>\n",
       "      <td>0.92</td>\n",
       "      <td>80.0</td>\n",
       "      <td>101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.67</td>\n",
       "      <td>70.0</td>\n",
       "      <td>99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.42</td>\n",
       "      <td>74.0</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                min   max  count\n",
       "Sex    Pclass                   \n",
       "female 1       2.00  63.0     85\n",
       "       2       2.00  57.0     74\n",
       "       3       0.75  63.0    102\n",
       "male   1       0.92  80.0    101\n",
       "       2       0.67  70.0     99\n",
       "       3       0.42  74.0    253"
      ]
     },
     "execution_count": 209,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby([\"Sex\", 'Pclass']).Age.agg(['min', 'max', 'count'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "透過 `unstack` 函式能讓你產生跟 `pivot_table` 函式相同的結果："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"3\" halign=\"left\">min</th>\n",
       "      <th colspan=\"3\" halign=\"left\">max</th>\n",
       "      <th colspan=\"3\" halign=\"left\">count</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Pclass</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Sex</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>female</th>\n",
       "      <td>2.00</td>\n",
       "      <td>2.00</td>\n",
       "      <td>0.75</td>\n",
       "      <td>63.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>63.0</td>\n",
       "      <td>85</td>\n",
       "      <td>74</td>\n",
       "      <td>102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>male</th>\n",
       "      <td>0.92</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.42</td>\n",
       "      <td>80.0</td>\n",
       "      <td>70.0</td>\n",
       "      <td>74.0</td>\n",
       "      <td>101</td>\n",
       "      <td>99</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         min               max             count         \n",
       "Pclass     1     2     3     1     2     3     1   2    3\n",
       "Sex                                                      \n",
       "female  2.00  2.00  0.75  63.0  57.0  63.0    85  74  102\n",
       "male    0.92  0.67  0.42  80.0  70.0  74.0   101  99  253"
      ]
     },
     "execution_count": 210,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby([\"Sex\", 'Pclass']).Age.agg(['min', 'max', 'count']).unstack()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "當然，你也可以直接使用 `pivot_table` 函式來匯總各組數據："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"3\" halign=\"left\">min</th>\n",
       "      <th colspan=\"3\" halign=\"left\">max</th>\n",
       "      <th colspan=\"3\" halign=\"left\">count</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Pclass</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Sex</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>female</th>\n",
       "      <td>2.00</td>\n",
       "      <td>2.00</td>\n",
       "      <td>0.75</td>\n",
       "      <td>63.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>63.0</td>\n",
       "      <td>85</td>\n",
       "      <td>74</td>\n",
       "      <td>102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>male</th>\n",
       "      <td>0.92</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.42</td>\n",
       "      <td>80.0</td>\n",
       "      <td>70.0</td>\n",
       "      <td>74.0</td>\n",
       "      <td>101</td>\n",
       "      <td>99</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         min               max             count         \n",
       "Pclass     1     2     3     1     2     3     1   2    3\n",
       "Sex                                                      \n",
       "female  2.00  2.00  0.75  63.0  57.0  63.0    85  74  102\n",
       "male    0.92  0.67  0.42  80.0  70.0  74.0   101  99  253"
      ]
     },
     "execution_count": 211,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.pivot_table(index='Sex', \n",
    "               columns='Pclass', \n",
    "               values='Age', \n",
    "               aggfunc=['min', 'max', 'count'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "依照背景不同，每個人會有偏好的 pandas 使用方式。選擇對你來說最直覺又好記的方式吧！"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 結合原始數據與匯總結果"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "不管是上節的 `groupby` 搭配 `agg` 還是 `pivot_table`，匯總結果都會以另外一個全新的 DataFrame 表示。\n",
    "\n",
    "有時候你會想直接把各組匯總的結果放到原本的 DataFrame 裡頭，方便比較原始樣本與匯總結果的差異。這時你可以使用 `transform` 函式："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_51aa6300_b014_11e9_8b6c_8c8590794fe2row0_col1 {\n",
       "            background-color:  rgb(153, 255, 51);\n",
       "        }    #T_51aa6300_b014_11e9_8b6c_8c8590794fe2row0_col8 {\n",
       "            background-color:  yellow;\n",
       "        }    #T_51aa6300_b014_11e9_8b6c_8c8590794fe2row0_col9 {\n",
       "            background-color:  rgb(153, 255, 51);\n",
       "        }    #T_51aa6300_b014_11e9_8b6c_8c8590794fe2row4_col1 {\n",
       "            background-color:  rgb(153, 255, 51);\n",
       "        }    #T_51aa6300_b014_11e9_8b6c_8c8590794fe2row4_col8 {\n",
       "            background-color:  yellow;\n",
       "        }    #T_51aa6300_b014_11e9_8b6c_8c8590794fe2row4_col9 {\n",
       "            background-color:  rgb(153, 255, 51);\n",
       "        }</style><table id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2\" ><thead>    <tr>        <th class=\"blank level0\" ></th>        <th class=\"col_heading level0 col0\" >Sex</th>        <th class=\"col_heading level0 col1\" >Age</th>        <th class=\"col_heading level0 col2\" >SibSp</th>        <th class=\"col_heading level0 col3\" >Parch</th>        <th class=\"col_heading level0 col4\" >Ticket</th>        <th class=\"col_heading level0 col5\" >Fare</th>        <th class=\"col_heading level0 col6\" >Cabin</th>        <th class=\"col_heading level0 col7\" >Embarked</th>        <th class=\"col_heading level0 col8\" >Avg_age</th>        <th class=\"col_heading level0 col9\" >Above_avg_age</th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row0_col0\" class=\"data row0 col0\" >male</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row0_col1\" class=\"data row0 col1\" >22</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row0_col2\" class=\"data row0 col2\" >1</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row0_col3\" class=\"data row0 col3\" >0</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row0_col4\" class=\"data row0 col4\" >A/5 21171</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row0_col5\" class=\"data row0 col5\" >7.25</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row0_col6\" class=\"data row0 col6\" >nan</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row0_col7\" class=\"data row0 col7\" >S</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row0_col8\" class=\"data row0 col8\" >30.7266</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row0_col9\" class=\"data row0 col9\" >no</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row1_col0\" class=\"data row1 col0\" >female</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row1_col1\" class=\"data row1 col1\" >38</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row1_col2\" class=\"data row1 col2\" >1</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row1_col3\" class=\"data row1 col3\" >0</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row1_col4\" class=\"data row1 col4\" >PC 17599</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row1_col5\" class=\"data row1 col5\" >71.2833</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row1_col6\" class=\"data row1 col6\" >C85</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row1_col7\" class=\"data row1 col7\" >C</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row1_col8\" class=\"data row1 col8\" >27.9157</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row1_col9\" class=\"data row1 col9\" >yes</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row2_col0\" class=\"data row2 col0\" >female</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row2_col1\" class=\"data row2 col1\" >26</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row2_col2\" class=\"data row2 col2\" >0</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row2_col3\" class=\"data row2 col3\" >0</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row2_col4\" class=\"data row2 col4\" >STON/O2. 3101282</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row2_col5\" class=\"data row2 col5\" >7.925</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row2_col6\" class=\"data row2 col6\" >nan</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row2_col7\" class=\"data row2 col7\" >S</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row2_col8\" class=\"data row2 col8\" >27.9157</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row2_col9\" class=\"data row2 col9\" >no</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row3_col0\" class=\"data row3 col0\" >female</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row3_col1\" class=\"data row3 col1\" >35</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row3_col2\" class=\"data row3 col2\" >1</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row3_col3\" class=\"data row3 col3\" >0</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row3_col4\" class=\"data row3 col4\" >113803</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row3_col5\" class=\"data row3 col5\" >53.1</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row3_col6\" class=\"data row3 col6\" >C123</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row3_col7\" class=\"data row3 col7\" >S</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row3_col8\" class=\"data row3 col8\" >27.9157</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row3_col9\" class=\"data row3 col9\" >yes</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row4_col0\" class=\"data row4 col0\" >male</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row4_col1\" class=\"data row4 col1\" >35</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row4_col2\" class=\"data row4 col2\" >0</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row4_col3\" class=\"data row4 col3\" >0</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row4_col4\" class=\"data row4 col4\" >373450</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row4_col5\" class=\"data row4 col5\" >8.05</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row4_col6\" class=\"data row4 col6\" >nan</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row4_col7\" class=\"data row4 col7\" >S</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row4_col8\" class=\"data row4 col8\" >30.7266</td>\n",
       "                        <td id=\"T_51aa6300_b014_11e9_8b6c_8c8590794fe2row4_col9\" class=\"data row4 col9\" >yes</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1a2e879cc0>"
      ]
     },
     "execution_count": 221,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df_titanic.copy()\n",
    "df['Avg_age'] = df.groupby(\"Sex\").Age.transform(\"mean\")\n",
    "df['Above_avg_age'] = df.apply(lambda x: 'yes' if x.Age > x.Avg_age else 'no', \n",
    "                               axis=1)\n",
    "# styling\n",
    "(df.loc[:4, 'Sex':]\n",
    " .style\n",
    " .highlight_max(subset=['Avg_age'])\n",
    " .applymap(lambda x: 'background-color: rgb(153, 255, 51)', \n",
    "           subset=pd.IndexSlice[[0, 4], ['Age', 'Above_avg_age']])\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "此例將所有乘客依照性別 `Sex` 分組之後，計算各組的平均年齡 `Age`，並利用 `transform` 函式將各組結果插入對應的乘客（列）裡頭。你會發現兩名男乘客跟平均男性壽命 `Avg_age` 欄位相比正好一上一下，這差異則反映到 `Above_avg_age` 欄位裡頭。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 對時間數據做匯總"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "給定一個跟時間相關的 DataFrame："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>C</th>\n",
       "      <th>D</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2000-03-31</th>\n",
       "      <td>-20.465499</td>\n",
       "      <td>-5.402825</td>\n",
       "      <td>4.790400</td>\n",
       "      <td>-6.949846</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-06-30</th>\n",
       "      <td>-3.573217</td>\n",
       "      <td>2.091491</td>\n",
       "      <td>12.725446</td>\n",
       "      <td>-10.215842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-09-30</th>\n",
       "      <td>-3.377268</td>\n",
       "      <td>3.600146</td>\n",
       "      <td>13.607341</td>\n",
       "      <td>6.309234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000-12-31</th>\n",
       "      <td>-11.181567</td>\n",
       "      <td>-16.471324</td>\n",
       "      <td>8.659892</td>\n",
       "      <td>5.619725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2001-03-31</th>\n",
       "      <td>7.148111</td>\n",
       "      <td>8.887097</td>\n",
       "      <td>-0.117279</td>\n",
       "      <td>-1.628799</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2001-06-30</th>\n",
       "      <td>9.765045</td>\n",
       "      <td>-1.160198</td>\n",
       "      <td>4.898406</td>\n",
       "      <td>12.054339</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2001-09-30</th>\n",
       "      <td>-8.067456</td>\n",
       "      <td>-9.746304</td>\n",
       "      <td>-12.500050</td>\n",
       "      <td>-8.970445</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2001-12-31</th>\n",
       "      <td>-3.300273</td>\n",
       "      <td>-11.148755</td>\n",
       "      <td>-13.194654</td>\n",
       "      <td>12.852938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2002-03-31</th>\n",
       "      <td>-6.421518</td>\n",
       "      <td>12.420474</td>\n",
       "      <td>-3.546766</td>\n",
       "      <td>20.694348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2002-06-30</th>\n",
       "      <td>11.967359</td>\n",
       "      <td>9.522958</td>\n",
       "      <td>1.557142</td>\n",
       "      <td>-2.778456</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    A          B          C          D\n",
       "2000-03-31 -20.465499  -5.402825   4.790400  -6.949846\n",
       "2000-06-30  -3.573217   2.091491  12.725446 -10.215842\n",
       "2000-09-30  -3.377268   3.600146  13.607341   6.309234\n",
       "2000-12-31 -11.181567 -16.471324   8.659892   5.619725\n",
       "2001-03-31   7.148111   8.887097  -0.117279  -1.628799\n",
       "2001-06-30   9.765045  -1.160198   4.898406  12.054339\n",
       "2001-09-30  -8.067456  -9.746304 -12.500050  -8.970445\n",
       "2001-12-31  -3.300273 -11.148755 -13.194654  12.852938\n",
       "2002-03-31  -6.421518  12.420474  -3.546766  20.694348\n",
       "2002-06-30  11.967359   9.522958   1.557142  -2.778456"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_date = pd.util.testing.makeTimeDataFrame(freq='Q').head(10) * 10\n",
    "df_date"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你可以利用 `resample` 函式來依照不同時間粒度匯總這個時間 DataFrame："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "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>year</th>\n",
       "      <th>A</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2000-12-31</td>\n",
       "      <td>-3.377268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2001-12-31</td>\n",
       "      <td>9.765045</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2002-12-31</td>\n",
       "      <td>11.967359</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        year          A\n",
       "0 2000-12-31  -3.377268\n",
       "1 2001-12-31   9.765045\n",
       "2 2002-12-31  11.967359"
      ]
     },
     "execution_count": 216,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(df_date.resample('Y').A.max()\n",
    " .reset_index()\n",
    " .rename({'index': 'year'}, axis=1)\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "此例中將不同年份（**Y**ear）的樣本分組，並從每一組的欄位 `A` 中選出最大值。你可以查看[官方 resampling 說明文件](https://pandas.pydata.org/pandas-docs/stable/user_guide/timeseries.html#resampling)來了解還有什麼時間粒度可以選擇（分鐘、小時、月份等等）。\n",
    "\n",
    "另外小細節是你可以利用 [numpy 的 broadcasting 運算](https://numpy.org/devdocs/user/theory.broadcasting.html)輕鬆地將 DataFrame 裡的所有數值做操作（初始 `df_date` 時的用到的 `* 10`）。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 簡易繪圖並修改預設樣式"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在 Python 世界裡有很多數據視覺化工具供你選擇，比方說經典的 [Matplotlib](https://matplotlib.org/) 以及在[淺談神經機器翻譯一文](https://leemeng.tw/neural-machine-translation-with-transformer-and-tensorflow2.html)中被我拿來視覺化矩陣運算的 [Seaborn](https://seaborn.pydata.org/)。你也可以使用前面看過的 [Chartify](https://leemeng.tw/chartify-a-simple-yet-powerful-python-data-visualization-tool-which-boost-your-productivity-as-a-data-scientist.html) 搭配 pandas 畫出美麗圖表。\n",
    "\n",
    "但有時，你只需要 pandas 內建的 `plot` 函式就能輕鬆地將一個 DataFrame 轉成統計圖：\n",
    "\n",
    "```python\n",
    "# 用 `plot` 一行畫圖。前兩行只是改變預設樣式\n",
    "import matplotlib.pyplot as plt\n",
    "plt.style.use(\"ggplot\")\n",
    "df.groupby(\"Pclass\").Survived.count().plot(\"barh\");\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!image\n",
    "- dark\n",
    "- pandas/pandas_plot.jpg"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "我們都是視覺動物，pandas 的 `plot` 函式讓你在進行[探索型數據分析（**E**xploratory **D**ata **A**nalysis, EDA）](https://en.wikipedia.org/wiki/Exploratory_data_analysis)、試著快速了解手上數據集時十分方便。我們在[淺談資料視覺化以及 ggplot2 實踐](https://leemeng.tw/data-visualization-from-matplotlib-to-ggplot2.html)一文也已經探討過類似概念：\n",
    "\n",
    "<br/>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!image\n",
    "- dark\n",
    "- ggplot2/data-vis-purpose.png\n",
    "- 在做 EDA 時只需要包含最重要且你自己能夠解讀的視覺變數，重點是效率\n",
    "- https://leemeng.tw/data-visualization-from-matplotlib-to-ggplot2.html"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br/>\n",
    "\n",
    "另外 pandas 底層預設使用 Matplotlib 繪圖，而用過 Matplotlib 的人都知道其初始的繪圖樣式實在不太討喜。你可以透過 `plt.style.available` 查看所有可供使用的繪圖樣式（style），並將喜歡的樣式透過 `plt.style.use()` 套用到所有 DataFrames 的 `plot` 函式："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['seaborn-dark',\n",
       " 'seaborn-darkgrid',\n",
       " 'seaborn-ticks',\n",
       " 'fivethirtyeight',\n",
       " 'seaborn-whitegrid',\n",
       " 'classic',\n",
       " '_classic_test',\n",
       " 'fast',\n",
       " 'seaborn-talk',\n",
       " 'seaborn-dark-palette',\n",
       " 'seaborn-bright',\n",
       " 'seaborn-pastel',\n",
       " 'grayscale',\n",
       " 'seaborn-notebook',\n",
       " 'ggplot',\n",
       " 'seaborn-colorblind',\n",
       " 'seaborn-muted',\n",
       " 'seaborn',\n",
       " 'Solarize_Light2',\n",
       " 'seaborn-paper',\n",
       " 'bmh',\n",
       " 'tableau-colorblind10',\n",
       " 'seaborn-white',\n",
       " 'dark_background',\n",
       " 'seaborn-poster',\n",
       " 'seaborn-deep']"
      ]
     },
     "execution_count": 206,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plt.style.available"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 與 pandas 相得益彰的實用工具"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "前面幾個章節介紹了不少 pandas 的使用技巧與操作概念，這節則介紹一些我認為十分適合跟 pandas 一起搭配使用的數據工具 / 函式庫。在說明每個工具的功能時，我都會使用你已經十分熟悉的鐵達尼號數據集作為範例 DataFrame："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "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>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>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     Sex   Age  SibSp  Parch  \\\n",
       "0            1         0       3    male  22.0      1      0   \n",
       "1            2         1       1  female  38.0      1      0   \n",
       "2            3         1       3  female  26.0      0      0   \n",
       "3            4         1       1  female  35.0      1      0   \n",
       "4            5         0       3    male  35.0      0      0   \n",
       "\n",
       "             Ticket     Fare Cabin Embarked  \n",
       "0         A/5 21171   7.2500   NaN        S  \n",
       "1          PC 17599  71.2833   C85        C  \n",
       "2  STON/O2. 3101282   7.9250   NaN        S  \n",
       "3            113803  53.1000  C123        S  \n",
       "4            373450   8.0500   NaN        S  "
      ]
     },
     "execution_count": 231,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df_titanic.copy()\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### [tqdm](https://github.com/tqdm/tqdm)：了解你的數據處理進度"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[tqdm](https://github.com/tqdm/tqdm) 是一個十分強大的 Python 進度條工具，且有[整合 pandas](https://github.com/tqdm/tqdm#pandas-integration)。此工具可以幫助我們了解 DataFrame `apply` 函式的進度。回想一下我們在[對某一軸套用相同運算](#對某一軸套用相同運算)一節做的一個簡單 `apply` 運算："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "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>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "      <th>存活</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>死亡</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "      <td>倖存</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>倖存</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "      <td>倖存</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>373450</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "      <td>死亡</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>330877</td>\n",
       "      <td>8.4583</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Q</td>\n",
       "      <td>死亡</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Survived  Pclass     Sex   Age  SibSp  Parch            Ticket     Fare  \\\n",
       "0         0       3    male  22.0      1      0         A/5 21171   7.2500   \n",
       "1         1       1  female  38.0      1      0          PC 17599  71.2833   \n",
       "2         1       3  female  26.0      0      0  STON/O2. 3101282   7.9250   \n",
       "3         1       1  female  35.0      1      0            113803  53.1000   \n",
       "4         0       3    male  35.0      0      0            373450   8.0500   \n",
       "5         0       3    male   NaN      0      0            330877   8.4583   \n",
       "\n",
       "  Cabin Embarked  存活  \n",
       "0   NaN        S  死亡  \n",
       "1   C85        C  倖存  \n",
       "2   NaN        S  倖存  \n",
       "3  C123        S  倖存  \n",
       "4   NaN        S  死亡  \n",
       "5   NaN        Q  死亡  "
      ]
     },
     "execution_count": 244,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['存活'] = df.Survived.apply(lambda x: '倖存' if x else '死亡')\n",
    "df.loc[:5, 'Survived':'存活']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在這個不到 1,000 筆的 DataFrame 做這樣的簡單運算不用一秒鐘，但實務上你可能常常需要對幾十萬、幾百萬筆數據分別做複雜的運算，這時候了解執行進度就是一件非常重要的事情。首先透過 [Anaconda](https://www.anaconda.com/) 安裝 tqdm：\n",
    "\n",
    "```bash\n",
    "conda install -c conda-forge tqdm \n",
    "```\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "如果你是在 Juypter 筆記本裡執行 pandas，可以用以下的方式將 tqdm 進度條整合到 `apply` 函式：\n",
    "\n",
    "```python\n",
    "from tqdm import tqdm_notebook\n",
    "tqdm_notebook().pandas()\n",
    "clear_output()\n",
    "\n",
    "# 只需將 `apply` 替換成 `progress_apply`\n",
    "df['存活'] = df.Survived.progress_apply(lambda x: '倖存' if x else '死亡')\n",
    "df.loc[:5, 'Survived':'存活']\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {},
   "outputs": [],
   "source": [
    "#ignore\n",
    "from tqdm import tqdm_notebook\n",
    "tqdm_notebook().pandas()\n",
    "df['存活'] = df.Survived.progress_apply(lambda x: '倖存' if x else '死亡')\n",
    "clear_output()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!image\n",
    "- dark\n",
    "- pandas/tqdm_progress_apply.jpg"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "透過使用 `progress_apply` 函式，我們可以得到跟使用 `apply` 函式一樣的結果，附贈進度條。相信我，在你 `apply` 的函式很複雜且樣本數很大時，你會很感謝有進度條的存在。你可以查看 [tqdm repo](https://github.com/tqdm/tqdm) 了解更多使用案例："
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!image\n",
    "- dark\n",
    "- pandas/tqdm-jupyter-1.gif"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### [swifter](https://github.com/jmcarpenter2/swifter)：加速你的數據處理"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[swifter](https://github.com/jmcarpenter2/swifter) 函式庫能以最有效率的方式執行 `apply` 函式。同樣可以使用 Anaconda 安裝：\n",
    "\n",
    "```bash\n",
    "conda install -c conda-forge swifter\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "接著讓我建立一個有 100 萬筆樣本的 DataFrame ，測試 swifter 與原版 `apply` 函式的效能差異："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "240 ms ± 5.24 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)\n",
      "5 ms ± 565 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)\n"
     ]
    }
   ],
   "source": [
    "import swifter\n",
    "df = pd.DataFrame(pd.np.random.rand(1000000, 1), columns=['x'])\n",
    "\n",
    "%timeit -n 10 df['x2'] = df['x'].apply(lambda x: x**2)\n",
    "%timeit -n 10 df['x2'] = df['x'].swifter.apply(lambda x: x**2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在這個簡單的平方運算中，swifter 版的 `apply` 函式在我的 mac 上的效率是原始 `apply` 函式的 48 倍。而要使用 swifter 你也只需要加上 `swifter` 關鍵字即可，何樂而不為？\n",
    "\n",
    "喔別擔心，剛剛的 tqdm 進度條仍然會繼續顯示：\n",
    "\n",
    "```python\n",
    "df = pd.DataFrame(pd.np.random.rand(100000, 1), columns=['x'])\n",
    "df['x2'] = df['x'].swifter.apply(lambda x: x**2)\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!image\n",
    "- dark\n",
    "- pandas/swifter_with_tqdm.png"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "魚與熊掌兩者皆得。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### [qgrid](https://github.com/quantopian/qgrid)：即時排序、篩選及編輯你的 DataFrame"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[qgrid](https://github.com/quantopian/qgrid) 是一個能讓你在 Jupyter 筆記本裡使用簡單 UI 瀏覽並操作 DataFrame 的 widget。一樣透過 Anaconda 安裝：\n",
    "\n",
    "```bash\n",
    "conda install -c tim_shawver qgrid\n",
    "jupyter nbextension enable --py --sys-prefix widgetsnbextension\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "實際的使用方式也非常簡單：\n",
    "\n",
    "```python\n",
    "import qgrid\n",
    "qgrid.set_grid_option('maxVisibleRows', 7)\n",
    "q = qgrid.show_grid(df_titanic)\n",
    "q\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!mp4\n",
    "- dark\n",
    "- images/pandas/qgrid_demo.mp4"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你可以簡單瀏覽 DataFrame 內容，也能依照欄位值篩選及排序樣本，甚至編輯裡頭的值。\n",
    "\n",
    "如果你想要取得當前狀態的 DataFrame，也能使用 qgrid 裡頭的 `get_changed_df` 函式：\n",
    "\n",
    "```python\n",
    "# 取得更改後的 DataFrame\n",
    "df_changed = q.get_changed_df()\n",
    "\n",
    "# styling\n",
    "(df_changed.iloc[:8, :]\n",
    " .style\n",
    " .applymap(lambda x: 'color:red',\n",
    "           subset=pd.IndexSlice[829, 'SibSp'])\n",
    ")\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!image\n",
    "- dark\n",
    "- pandas/qgrid_get_changed_df.png"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "注意我剛剛在 qgrid 裡頭將 `0` 改成 `2`，而這個變動被反映到新的 DataFrame `df_changed` 裡頭。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### [pandas-profiling](https://github.com/pandas-profiling/pandas-profiling)：你的一鍵 EDA 神器"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "廢話不多說，先附上安裝指令：\n",
    "\n",
    "```bash\n",
    "conda install -c conda-forge pandas-profiling\n",
    "```\n",
    "\n",
    "當碰到新的資料集時，你常會需要做 EDA 來探索並了解手上的數據。常見的基礎步驟有：\n",
    "- 隨機抽樣 DataFrame 並用肉眼觀察樣本\n",
    "- 了解樣本數有多少、各類型的變數有多少\n",
    "- 了解 DataFrame 裡有多少空值\n",
    "- 看數值欄位的分佈，欄位之間的相關係數\n",
    "- 看分類型欄位的有多少不同的值，分佈為何\n",
    "\n",
    "等等。這些在你熟悉本文的 pandas 技巧以後都應該跟吃飯喝水一樣，但很多時候就是例行公事。\n",
    "\n",
    "這時你可以使用 [pandas-profiling](https://github.com/pandas-profiling/pandas-profiling) 讓它自動產生簡單報表，總結 DataFrame 裡的數據性質：\n",
    "\n",
    "```python\n",
    "import pandas_profiling\n",
    "df = df_titanic.copy()\n",
    "\n",
    "# 一行報表：將想觀察的 DataFrame 丟進去就完工了\n",
    "pandas_profiling.ProfileReport(df)\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!mp4\n",
    "- dark\n",
    "- images/pandas/pandas_profiling.mp4"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "有了自動生成的數據分析報表，你不再需要自己寫繁瑣的程式碼就能「健檢」手上的數據集。這讓你能對數據更快地展開 domain-specific 的研究與分析，大大提升效率。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 結語：實際運用所學"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "呼！本文的 pandas 旅程到此告一段落啦！\n",
    "\n",
    "我想在其他地方你應該是找不到跟本文一樣囉哩八唆的 pandas 教學文章了。本文雖長，但涵蓋的都是我認為十分實用的 pandas 使用手法，希望你有從中學到些東西，並開始自己的數據處理與分析之旅。如果你之後想要複習，可以點開頁面左側的章節傳送門，測試自己還記得多少技巧。\n",
    "\n",
    "接下來最重要的是培養你自己的 pandas **肌肉記憶**：重複應用你在本文學到的東西，分析自己感興趣的任何數據並內化這些知識。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!image\n",
    "- pandas/danial-ricaros-FCHlYvR5gJI-unsplash.jpg\n",
    "- Now it's your show time"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "pandas 是一個非常強大的資料處理工具，本文雖長，仍無法涵蓋所有東西。不過你現在已經學會作為一個資料科學家所需要的所有基礎 pandas 手法。除了多加練習找手感以外，我能給你的最後一個建議就是在 Jupyter 筆記本裡頭多加利用 `Shift` + `tab` 查看 pandas 函式的 documentation，很多時候你會有新的發現。\n",
    "\n",
    "如果你有任何其他 pandas 技巧，也請不吝留言與我分享！我懂的技巧不多，而現在輪到你教我了：）"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#ignore\n",
    "- output\n",
    "    - windows friendly output: `to_csv(encoding=\"cp932\")`\n",
    "- powerful tools\n",
    "    - Facets\n",
    "        - https://www.evernote.com/l/AETaGMqtguRAKbEF9z4hWtF9HazEVkWr70c\n",
    "    - cufflinks and plotly\n",
    "        - https://www.kdnuggets.com/2019/07/10-simple-hacks-speed-data-analysis-python.html\n",
    "\n",
    "- post-release\n",
    "    - 修改 chartify 連結，連回 pandsa 文章\n",
    "    - 檢查 pandas 文內 chartify 以及 ggplot 的連結對不對\n"
   ]
  }
 ],
 "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": "3.8.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}