Chapter-10/10.2 Data Aggregation（数据聚合）.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 10.2 Data Aggregation（数据聚合）\n",
    "\n",
    "聚合（Aggregation）指的是一些数据转化（data transformation），这些数据转化能从数组中产生标量（scalar values）。下面的例子就是一些聚合方法，包括mean, count, min and sum。我们可能会好奇，在一个GroupBy对象上调用mean()的时候，究竟发生了什么。一些常见的聚合，比如下表，实现方法上都已经被优化过了。当然，我们可以使用的聚合方法不止这些：\n",
    "\n",
    "![](http://oydgk2hgw.bkt.clouddn.com/pydata-book/sugsj.png)\n",
    "\n",
    "我们可以使用自己设计的聚合方法，而且可以调用分组后对象上的任意方法。例如，我们可以调用quantile来计算Series或DataFrame中列的样本的百分数。\n",
    "\n",
    "尽管quantile并不是专门为GroupBy对象设计的方法，这是一个Series方法，但仍可以被GroupBy对象使用。**GroupBy会对Series进行切片（slice up），并对于切片后的每一部分调用piece.quantile(0.9)，然后把每部分的结果整合到一起**：\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>data1</th>\n",
       "      <th>data2</th>\n",
       "      <th>key1</th>\n",
       "      <th>key2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.707738</td>\n",
       "      <td>0.186729</td>\n",
       "      <td>a</td>\n",
       "      <td>one</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.069831</td>\n",
       "      <td>1.305796</td>\n",
       "      <td>a</td>\n",
       "      <td>two</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-2.291339</td>\n",
       "      <td>-1.609071</td>\n",
       "      <td>b</td>\n",
       "      <td>one</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.348090</td>\n",
       "      <td>-0.294999</td>\n",
       "      <td>b</td>\n",
       "      <td>two</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.341176</td>\n",
       "      <td>0.429461</td>\n",
       "      <td>a</td>\n",
       "      <td>one</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      data1     data2 key1 key2\n",
       "0  1.707738  0.186729    a  one\n",
       "1  1.069831  1.305796    a  two\n",
       "2 -2.291339 -1.609071    b  one\n",
       "3  1.348090 -0.294999    b  two\n",
       "4  0.341176  0.429461    a  one"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame({'key1' : ['a', 'a', 'b', 'b', 'a'],\n",
    "                   'key2' : ['one', 'two', 'one', 'two', 'one'], \n",
    "                   'data1' : np.random.randn(5), \n",
    "                   'data2' : np.random.randn(5)})\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "a\n",
      "      data1     data2 key1 key2\n",
      "0  1.707738  0.186729    a  one\n",
      "1  1.069831  1.305796    a  two\n",
      "4  0.341176  0.429461    a  one\n",
      "b\n",
      "      data1     data2 key1 key2\n",
      "2 -2.291339 -1.609071    b  one\n",
      "3  1.348090 -0.294999    b  two\n"
     ]
    }
   ],
   "source": [
    "grouped = df.groupby('key1')\n",
    "for key, group in grouped:\n",
    "    print(key)\n",
    "    print(group)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "key1\n",
       "a    1.580157\n",
       "b    0.984147\n",
       "Name: data1, dtype: float64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grouped['data1'].quantile(0.9)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "如果想用自己设计的聚合函数，把用于聚合数组的函数传入到aggregate或agg方法即可："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def peak_to_peak(arr):\n",
    "    return arr.max() - arr.min()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>data1</th>\n",
       "      <th>data2</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>key1</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>1.366563</td>\n",
       "      <td>1.119067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>3.639430</td>\n",
       "      <td>1.314072</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         data1     data2\n",
       "key1                    \n",
       "a     1.366563  1.119067\n",
       "b     3.639430  1.314072"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grouped.agg(peak_to_peak)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "我们发现很多方法，比如describe，也能正常使用，尽管严格的来说，这并不是聚合："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>data1</th>\n",
       "      <th>data2</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>key1</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"8\" valign=\"top\">a</th>\n",
       "      <th>count</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1.039582</td>\n",
       "      <td>0.640662</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.683783</td>\n",
       "      <td>0.588670</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.341176</td>\n",
       "      <td>0.186729</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.705503</td>\n",
       "      <td>0.308095</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.069831</td>\n",
       "      <td>0.429461</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.388785</td>\n",
       "      <td>0.867629</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.707738</td>\n",
       "      <td>1.305796</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"8\" valign=\"top\">b</th>\n",
       "      <th>count</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>-0.471624</td>\n",
       "      <td>-0.952035</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>2.573465</td>\n",
       "      <td>0.929189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-2.291339</td>\n",
       "      <td>-1.609071</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-1.381482</td>\n",
       "      <td>-1.280553</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>-0.471624</td>\n",
       "      <td>-0.952035</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.438233</td>\n",
       "      <td>-0.623517</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.348090</td>\n",
       "      <td>-0.294999</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               data1     data2\n",
       "key1                          \n",
       "a    count  3.000000  3.000000\n",
       "     mean   1.039582  0.640662\n",
       "     std    0.683783  0.588670\n",
       "     min    0.341176  0.186729\n",
       "     25%    0.705503  0.308095\n",
       "     50%    1.069831  0.429461\n",
       "     75%    1.388785  0.867629\n",
       "     max    1.707738  1.305796\n",
       "b    count  2.000000  2.000000\n",
       "     mean  -0.471624 -0.952035\n",
       "     std    2.573465  0.929189\n",
       "     min   -2.291339 -1.609071\n",
       "     25%   -1.381482 -1.280553\n",
       "     50%   -0.471624 -0.952035\n",
       "     75%    0.438233 -0.623517\n",
       "     max    1.348090 -0.294999"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grouped.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "细节的部分在10.3会进行更多解释。\n",
    "\n",
    "注意：自定义的函数会比上面表中的函数慢一些，上面的函数时优化过的，而自定义的函数会有一些额外的计算，所以慢一些。\n",
    "\n",
    "# 1 Column-Wise and Multiple Function Application（列对列和多函数应用）\n",
    "\n",
    "让我们回到tipping数据集。加载数据及后，我们添加一列用于描述小费的百分比："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "tips = pd.read_csv('../examples/tips.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Add tip percentage of total bill\n",
    "tips['tip_pct'] = tips['tip'] / tips['total_bill']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_bill</th>\n",
       "      <th>tip</th>\n",
       "      <th>smoker</th>\n",
       "      <th>day</th>\n",
       "      <th>time</th>\n",
       "      <th>size</th>\n",
       "      <th>tip_pct</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>16.99</td>\n",
       "      <td>1.01</td>\n",
       "      <td>No</td>\n",
       "      <td>Sun</td>\n",
       "      <td>Dinner</td>\n",
       "      <td>2</td>\n",
       "      <td>0.059447</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10.34</td>\n",
       "      <td>1.66</td>\n",
       "      <td>No</td>\n",
       "      <td>Sun</td>\n",
       "      <td>Dinner</td>\n",
       "      <td>3</td>\n",
       "      <td>0.160542</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>21.01</td>\n",
       "      <td>3.50</td>\n",
       "      <td>No</td>\n",
       "      <td>Sun</td>\n",
       "      <td>Dinner</td>\n",
       "      <td>3</td>\n",
       "      <td>0.166587</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>23.68</td>\n",
       "      <td>3.31</td>\n",
       "      <td>No</td>\n",
       "      <td>Sun</td>\n",
       "      <td>Dinner</td>\n",
       "      <td>2</td>\n",
       "      <td>0.139780</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>24.59</td>\n",
       "      <td>3.61</td>\n",
       "      <td>No</td>\n",
       "      <td>Sun</td>\n",
       "      <td>Dinner</td>\n",
       "      <td>4</td>\n",
       "      <td>0.146808</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>25.29</td>\n",
       "      <td>4.71</td>\n",
       "      <td>No</td>\n",
       "      <td>Sun</td>\n",
       "      <td>Dinner</td>\n",
       "      <td>4</td>\n",
       "      <td>0.186240</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   total_bill   tip smoker  day    time  size   tip_pct\n",
       "0       16.99  1.01     No  Sun  Dinner     2  0.059447\n",
       "1       10.34  1.66     No  Sun  Dinner     3  0.160542\n",
       "2       21.01  3.50     No  Sun  Dinner     3  0.166587\n",
       "3       23.68  3.31     No  Sun  Dinner     2  0.139780\n",
       "4       24.59  3.61     No  Sun  Dinner     4  0.146808\n",
       "5       25.29  4.71     No  Sun  Dinner     4  0.186240"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tips[:6]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "我们可以看到，对series或DataFrame进行聚合，其实就是通过aggregate使用合适的函数，或者调用一些像mean或std这样的方法。然而，我们可能想要在列上使用不同的函数进行聚合，又或者想要一次执行多个函数。幸运的是，这是可能的，下面将通过一些例子来说明。首先，对于tips数据集，先用day和smoker进行分组："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "grouped = tips.groupby(['day', 'smoker'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "对于像是上面表格10-1中的一些描述性统计，我们可以直接传入函数的名字，即字符串："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "grouped_pct = grouped['tip_pct']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('Fri', 'No')\n",
      "91    0.155625\n",
      "94    0.142857\n",
      "Name: tip_pct, dtype: float64 \n",
      "\n",
      "('Fri', 'Yes')\n",
      "90    0.103555\n",
      "92    0.173913\n",
      "Name: tip_pct, dtype: float64 \n",
      "\n",
      "('Sat', 'No')\n",
      "19    0.162228\n",
      "20    0.227679\n",
      "Name: tip_pct, dtype: float64 \n",
      "\n",
      "('Sat', 'Yes')\n",
      "56    0.078927\n",
      "58    0.156584\n",
      "Name: tip_pct, dtype: float64 \n",
      "\n",
      "('Sun', 'No')\n",
      "0    0.059447\n",
      "1    0.160542\n",
      "Name: tip_pct, dtype: float64 \n",
      "\n",
      "('Sun', 'Yes')\n",
      "164    0.171331\n",
      "172    0.710345\n",
      "Name: tip_pct, dtype: float64 \n",
      "\n",
      "('Thur', 'No')\n",
      "77    0.147059\n",
      "78    0.131810\n",
      "Name: tip_pct, dtype: float64 \n",
      "\n",
      "('Thur', 'Yes')\n",
      "80    0.154321\n",
      "83    0.152999\n",
      "Name: tip_pct, dtype: float64 \n",
      "\n"
     ]
    }
   ],
   "source": [
    "for name, group in grouped_pct:\n",
    "    print(name)\n",
    "    print(group[:2], '\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "day   smoker\n",
       "Fri   No        0.151650\n",
       "      Yes       0.174783\n",
       "Sat   No        0.158048\n",
       "      Yes       0.147906\n",
       "Sun   No        0.160113\n",
       "      Yes       0.187250\n",
       "Thur  No        0.160298\n",
       "      Yes       0.163863\n",
       "Name: tip_pct, dtype: float64"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grouped_pct.agg('mean')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "如果我们把函数或函数的名字作为一个list传入，我们会得到一个DataFrame，每列的名字就是函数的名字："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>peak_to_peak</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th>smoker</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Fri</th>\n",
       "      <th>No</th>\n",
       "      <td>0.151650</td>\n",
       "      <td>0.028123</td>\n",
       "      <td>0.067349</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.174783</td>\n",
       "      <td>0.051293</td>\n",
       "      <td>0.159925</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Sat</th>\n",
       "      <th>No</th>\n",
       "      <td>0.158048</td>\n",
       "      <td>0.039767</td>\n",
       "      <td>0.235193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.147906</td>\n",
       "      <td>0.061375</td>\n",
       "      <td>0.290095</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Sun</th>\n",
       "      <th>No</th>\n",
       "      <td>0.160113</td>\n",
       "      <td>0.042347</td>\n",
       "      <td>0.193226</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.187250</td>\n",
       "      <td>0.154134</td>\n",
       "      <td>0.644685</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Thur</th>\n",
       "      <th>No</th>\n",
       "      <td>0.160298</td>\n",
       "      <td>0.038774</td>\n",
       "      <td>0.193350</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.163863</td>\n",
       "      <td>0.039389</td>\n",
       "      <td>0.151240</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 mean       std  peak_to_peak\n",
       "day  smoker                                  \n",
       "Fri  No      0.151650  0.028123      0.067349\n",
       "     Yes     0.174783  0.051293      0.159925\n",
       "Sat  No      0.158048  0.039767      0.235193\n",
       "     Yes     0.147906  0.061375      0.290095\n",
       "Sun  No      0.160113  0.042347      0.193226\n",
       "     Yes     0.187250  0.154134      0.644685\n",
       "Thur No      0.160298  0.038774      0.193350\n",
       "     Yes     0.163863  0.039389      0.151240"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# def peak_to_peak(arr):\n",
    "#     return arr.max() - arr.min()\n",
    "grouped_pct.agg(['mean', 'std', peak_to_peak])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "上面我们把多个聚合函数作为一个list传入给agg，这些函数会独立对每一个组进行计算。\n",
    "\n",
    "上面结果的列名是自动给出的，当然，我们也可以更改这些列名。这种情况下，传入一个由tuple组成的list，每个tuple的格式是`(name, function)`，每个元组的第一个元素会被用于作为DataFrame的列名（我们可以认为这个二元元组list是一个有序的映射）："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>foo</th>\n",
       "      <th>bar</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th>smoker</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Fri</th>\n",
       "      <th>No</th>\n",
       "      <td>0.151650</td>\n",
       "      <td>0.028123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.174783</td>\n",
       "      <td>0.051293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Sat</th>\n",
       "      <th>No</th>\n",
       "      <td>0.158048</td>\n",
       "      <td>0.039767</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.147906</td>\n",
       "      <td>0.061375</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Sun</th>\n",
       "      <th>No</th>\n",
       "      <td>0.160113</td>\n",
       "      <td>0.042347</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.187250</td>\n",
       "      <td>0.154134</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Thur</th>\n",
       "      <th>No</th>\n",
       "      <td>0.160298</td>\n",
       "      <td>0.038774</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.163863</td>\n",
       "      <td>0.039389</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  foo       bar\n",
       "day  smoker                    \n",
       "Fri  No      0.151650  0.028123\n",
       "     Yes     0.174783  0.051293\n",
       "Sat  No      0.158048  0.039767\n",
       "     Yes     0.147906  0.061375\n",
       "Sun  No      0.160113  0.042347\n",
       "     Yes     0.187250  0.154134\n",
       "Thur No      0.160298  0.038774\n",
       "     Yes     0.163863  0.039389"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grouped_pct.agg([('foo', 'mean'), ('bar', np.std)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "如果是处理一个DataFrame，我们有更多的选择，我们可以用一个含有多个函数的list应用到所有的列上，也可以在不同的列上应用不同的函数。演示一下，假设我们想要在tip_pct和total_bill这两列上，计算三个相同的统计指标："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "functions = ['count', 'mean', 'max']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th colspan=\"3\" halign=\"left\">tip_pct</th>\n",
       "      <th colspan=\"3\" halign=\"left\">total_bill</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>max</th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th>smoker</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 rowspan=\"2\" valign=\"top\">Fri</th>\n",
       "      <th>No</th>\n",
       "      <td>4</td>\n",
       "      <td>0.151650</td>\n",
       "      <td>0.187735</td>\n",
       "      <td>4</td>\n",
       "      <td>18.420000</td>\n",
       "      <td>22.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>15</td>\n",
       "      <td>0.174783</td>\n",
       "      <td>0.263480</td>\n",
       "      <td>15</td>\n",
       "      <td>16.813333</td>\n",
       "      <td>40.17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Sat</th>\n",
       "      <th>No</th>\n",
       "      <td>45</td>\n",
       "      <td>0.158048</td>\n",
       "      <td>0.291990</td>\n",
       "      <td>45</td>\n",
       "      <td>19.661778</td>\n",
       "      <td>48.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>42</td>\n",
       "      <td>0.147906</td>\n",
       "      <td>0.325733</td>\n",
       "      <td>42</td>\n",
       "      <td>21.276667</td>\n",
       "      <td>50.81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Sun</th>\n",
       "      <th>No</th>\n",
       "      <td>57</td>\n",
       "      <td>0.160113</td>\n",
       "      <td>0.252672</td>\n",
       "      <td>57</td>\n",
       "      <td>20.506667</td>\n",
       "      <td>48.17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>19</td>\n",
       "      <td>0.187250</td>\n",
       "      <td>0.710345</td>\n",
       "      <td>19</td>\n",
       "      <td>24.120000</td>\n",
       "      <td>45.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Thur</th>\n",
       "      <th>No</th>\n",
       "      <td>45</td>\n",
       "      <td>0.160298</td>\n",
       "      <td>0.266312</td>\n",
       "      <td>45</td>\n",
       "      <td>17.113111</td>\n",
       "      <td>41.19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>17</td>\n",
       "      <td>0.163863</td>\n",
       "      <td>0.241255</td>\n",
       "      <td>17</td>\n",
       "      <td>19.190588</td>\n",
       "      <td>43.11</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            tip_pct                     total_bill                  \n",
       "              count      mean       max      count       mean    max\n",
       "day  smoker                                                         \n",
       "Fri  No           4  0.151650  0.187735          4  18.420000  22.75\n",
       "     Yes         15  0.174783  0.263480         15  16.813333  40.17\n",
       "Sat  No          45  0.158048  0.291990         45  19.661778  48.33\n",
       "     Yes         42  0.147906  0.325733         42  21.276667  50.81\n",
       "Sun  No          57  0.160113  0.252672         57  20.506667  48.17\n",
       "     Yes         19  0.187250  0.710345         19  24.120000  45.35\n",
       "Thur No          45  0.160298  0.266312         45  17.113111  41.19\n",
       "     Yes         17  0.163863  0.241255         17  19.190588  43.11"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result = grouped['tip_pct', 'total_bill'].agg(functions)\n",
    "result"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "我们可以看到，结果中的DataFrame有多层级的列（hierarchical columns）。另外一种做法有相同的效果，即我们对于每一列单独进行聚合（aggregating each column separately），然后使用concat把结果都结合在一起，然后用列名作为keys参数："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th>smoker</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Fri</th>\n",
       "      <th>No</th>\n",
       "      <td>4</td>\n",
       "      <td>0.151650</td>\n",
       "      <td>0.187735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>15</td>\n",
       "      <td>0.174783</td>\n",
       "      <td>0.263480</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Sat</th>\n",
       "      <th>No</th>\n",
       "      <td>45</td>\n",
       "      <td>0.158048</td>\n",
       "      <td>0.291990</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>42</td>\n",
       "      <td>0.147906</td>\n",
       "      <td>0.325733</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Sun</th>\n",
       "      <th>No</th>\n",
       "      <td>57</td>\n",
       "      <td>0.160113</td>\n",
       "      <td>0.252672</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>19</td>\n",
       "      <td>0.187250</td>\n",
       "      <td>0.710345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Thur</th>\n",
       "      <th>No</th>\n",
       "      <td>45</td>\n",
       "      <td>0.160298</td>\n",
       "      <td>0.266312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>17</td>\n",
       "      <td>0.163863</td>\n",
       "      <td>0.241255</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             count      mean       max\n",
       "day  smoker                           \n",
       "Fri  No          4  0.151650  0.187735\n",
       "     Yes        15  0.174783  0.263480\n",
       "Sat  No         45  0.158048  0.291990\n",
       "     Yes        42  0.147906  0.325733\n",
       "Sun  No         57  0.160113  0.252672\n",
       "     Yes        19  0.187250  0.710345\n",
       "Thur No         45  0.160298  0.266312\n",
       "     Yes        17  0.163863  0.241255"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result['tip_pct']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "我们之前提到过，可以用元组组成的list来自己定义列名："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "ftuples = [('Durchschnitt', 'mean'), ('Abweichung', np.var)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">tip_pct</th>\n",
       "      <th colspan=\"2\" halign=\"left\">total_bill</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>Durchschnitt</th>\n",
       "      <th>Abweichung</th>\n",
       "      <th>Durchschnitt</th>\n",
       "      <th>Abweichung</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th>smoker</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Fri</th>\n",
       "      <th>No</th>\n",
       "      <td>0.151650</td>\n",
       "      <td>0.000791</td>\n",
       "      <td>18.420000</td>\n",
       "      <td>25.596333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.174783</td>\n",
       "      <td>0.002631</td>\n",
       "      <td>16.813333</td>\n",
       "      <td>82.562438</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Sat</th>\n",
       "      <th>No</th>\n",
       "      <td>0.158048</td>\n",
       "      <td>0.001581</td>\n",
       "      <td>19.661778</td>\n",
       "      <td>79.908965</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.147906</td>\n",
       "      <td>0.003767</td>\n",
       "      <td>21.276667</td>\n",
       "      <td>101.387535</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Sun</th>\n",
       "      <th>No</th>\n",
       "      <td>0.160113</td>\n",
       "      <td>0.001793</td>\n",
       "      <td>20.506667</td>\n",
       "      <td>66.099980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.187250</td>\n",
       "      <td>0.023757</td>\n",
       "      <td>24.120000</td>\n",
       "      <td>109.046044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Thur</th>\n",
       "      <th>No</th>\n",
       "      <td>0.160298</td>\n",
       "      <td>0.001503</td>\n",
       "      <td>17.113111</td>\n",
       "      <td>59.625081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.163863</td>\n",
       "      <td>0.001551</td>\n",
       "      <td>19.190588</td>\n",
       "      <td>69.808518</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 tip_pct              total_bill            \n",
       "            Durchschnitt Abweichung Durchschnitt  Abweichung\n",
       "day  smoker                                                 \n",
       "Fri  No         0.151650   0.000791    18.420000   25.596333\n",
       "     Yes        0.174783   0.002631    16.813333   82.562438\n",
       "Sat  No         0.158048   0.001581    19.661778   79.908965\n",
       "     Yes        0.147906   0.003767    21.276667  101.387535\n",
       "Sun  No         0.160113   0.001793    20.506667   66.099980\n",
       "     Yes        0.187250   0.023757    24.120000  109.046044\n",
       "Thur No         0.160298   0.001503    17.113111   59.625081\n",
       "     Yes        0.163863   0.001551    19.190588   69.808518"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grouped['tip_pct', 'total_bill'].agg(ftuples)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "现在，假设我们想要把不同的函数用到一列或多列上。要做到这一点，给agg传递一个dict，这个dict需要包含映射关系，用来表示列名和函数之间的对应关系："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>tip</th>\n",
       "      <th>size</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th>smoker</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Fri</th>\n",
       "      <th>No</th>\n",
       "      <td>3.50</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>4.73</td>\n",
       "      <td>31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Sat</th>\n",
       "      <th>No</th>\n",
       "      <td>9.00</td>\n",
       "      <td>115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>10.00</td>\n",
       "      <td>104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Sun</th>\n",
       "      <th>No</th>\n",
       "      <td>6.00</td>\n",
       "      <td>167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>6.50</td>\n",
       "      <td>49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Thur</th>\n",
       "      <th>No</th>\n",
       "      <td>6.70</td>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>5.00</td>\n",
       "      <td>40</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               tip  size\n",
       "day  smoker             \n",
       "Fri  No       3.50     9\n",
       "     Yes      4.73    31\n",
       "Sat  No       9.00   115\n",
       "     Yes     10.00   104\n",
       "Sun  No       6.00   167\n",
       "     Yes      6.50    49\n",
       "Thur No       6.70   112\n",
       "     Yes      5.00    40"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grouped.agg({'tip': np.max, 'size': 'sum'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th colspan=\"4\" halign=\"left\">tip_pct</th>\n",
       "      <th>size</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>min</th>\n",
       "      <th>max</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>sum</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th>smoker</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 rowspan=\"2\" valign=\"top\">Fri</th>\n",
       "      <th>No</th>\n",
       "      <td>0.120385</td>\n",
       "      <td>0.187735</td>\n",
       "      <td>0.151650</td>\n",
       "      <td>0.028123</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.103555</td>\n",
       "      <td>0.263480</td>\n",
       "      <td>0.174783</td>\n",
       "      <td>0.051293</td>\n",
       "      <td>31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Sat</th>\n",
       "      <th>No</th>\n",
       "      <td>0.056797</td>\n",
       "      <td>0.291990</td>\n",
       "      <td>0.158048</td>\n",
       "      <td>0.039767</td>\n",
       "      <td>115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.035638</td>\n",
       "      <td>0.325733</td>\n",
       "      <td>0.147906</td>\n",
       "      <td>0.061375</td>\n",
       "      <td>104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Sun</th>\n",
       "      <th>No</th>\n",
       "      <td>0.059447</td>\n",
       "      <td>0.252672</td>\n",
       "      <td>0.160113</td>\n",
       "      <td>0.042347</td>\n",
       "      <td>167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.065660</td>\n",
       "      <td>0.710345</td>\n",
       "      <td>0.187250</td>\n",
       "      <td>0.154134</td>\n",
       "      <td>49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Thur</th>\n",
       "      <th>No</th>\n",
       "      <td>0.072961</td>\n",
       "      <td>0.266312</td>\n",
       "      <td>0.160298</td>\n",
       "      <td>0.038774</td>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>0.090014</td>\n",
       "      <td>0.241255</td>\n",
       "      <td>0.163863</td>\n",
       "      <td>0.039389</td>\n",
       "      <td>40</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              tip_pct                               size\n",
       "                  min       max      mean       std  sum\n",
       "day  smoker                                             \n",
       "Fri  No      0.120385  0.187735  0.151650  0.028123    9\n",
       "     Yes     0.103555  0.263480  0.174783  0.051293   31\n",
       "Sat  No      0.056797  0.291990  0.158048  0.039767  115\n",
       "     Yes     0.035638  0.325733  0.147906  0.061375  104\n",
       "Sun  No      0.059447  0.252672  0.160113  0.042347  167\n",
       "     Yes     0.065660  0.710345  0.187250  0.154134   49\n",
       "Thur No      0.072961  0.266312  0.160298  0.038774  112\n",
       "     Yes     0.090014  0.241255  0.163863  0.039389   40"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grouped.agg({'tip_pct': ['min', 'max', 'mean', 'std'],\n",
    "             'size': 'sum'})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "只有当多个函数用于至少一列的时候，DataFrame才会有多层级列（hierarchical columns）\n",
    "\n",
    "# 2 Returning Aggregated Data Without Row Indexes（不使用行索引返回聚合数据）\n",
    "\n",
    "目前为止提到的所有例子，最后返回的聚合数据都是有索引的，而且这个索引默认是多层级索引，这个索引是由不同的组键的组合构成的（unique group key combinations）。因为我们并不是总需要返回这种索引，所以我们可以取消这种模式，在调用groupby的时候设定as_index=False即可："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>day</th>\n",
       "      <th>smoker</th>\n",
       "      <th>total_bill</th>\n",
       "      <th>tip</th>\n",
       "      <th>size</th>\n",
       "      <th>tip_pct</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Fri</td>\n",
       "      <td>No</td>\n",
       "      <td>18.420000</td>\n",
       "      <td>2.812500</td>\n",
       "      <td>2.250000</td>\n",
       "      <td>0.151650</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Fri</td>\n",
       "      <td>Yes</td>\n",
       "      <td>16.813333</td>\n",
       "      <td>2.714000</td>\n",
       "      <td>2.066667</td>\n",
       "      <td>0.174783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Sat</td>\n",
       "      <td>No</td>\n",
       "      <td>19.661778</td>\n",
       "      <td>3.102889</td>\n",
       "      <td>2.555556</td>\n",
       "      <td>0.158048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Sat</td>\n",
       "      <td>Yes</td>\n",
       "      <td>21.276667</td>\n",
       "      <td>2.875476</td>\n",
       "      <td>2.476190</td>\n",
       "      <td>0.147906</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Sun</td>\n",
       "      <td>No</td>\n",
       "      <td>20.506667</td>\n",
       "      <td>3.167895</td>\n",
       "      <td>2.929825</td>\n",
       "      <td>0.160113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Sun</td>\n",
       "      <td>Yes</td>\n",
       "      <td>24.120000</td>\n",
       "      <td>3.516842</td>\n",
       "      <td>2.578947</td>\n",
       "      <td>0.187250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Thur</td>\n",
       "      <td>No</td>\n",
       "      <td>17.113111</td>\n",
       "      <td>2.673778</td>\n",
       "      <td>2.488889</td>\n",
       "      <td>0.160298</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Thur</td>\n",
       "      <td>Yes</td>\n",
       "      <td>19.190588</td>\n",
       "      <td>3.030000</td>\n",
       "      <td>2.352941</td>\n",
       "      <td>0.163863</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    day smoker  total_bill       tip      size   tip_pct\n",
       "0   Fri     No   18.420000  2.812500  2.250000  0.151650\n",
       "1   Fri    Yes   16.813333  2.714000  2.066667  0.174783\n",
       "2   Sat     No   19.661778  3.102889  2.555556  0.158048\n",
       "3   Sat    Yes   21.276667  2.875476  2.476190  0.147906\n",
       "4   Sun     No   20.506667  3.167895  2.929825  0.160113\n",
       "5   Sun    Yes   24.120000  3.516842  2.578947  0.187250\n",
       "6  Thur     No   17.113111  2.673778  2.488889  0.160298\n",
       "7  Thur    Yes   19.190588  3.030000  2.352941  0.163863"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tips.groupby(['day', 'smoker'], as_index=False).mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "当然，我们也可以在上面的结果上直接调用reset_index，这样的话就能得到之前那种多层级索引的结果。不过使用as_index=False方法可以避免一些不必要的计算。"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [py35]",
   "language": "python",
   "name": "Python [py35]"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}