从 DataFrame 的行或列中抽取一个 Series 或一系列值的单个值（如总和或平均值）。与 Numpy 数组中类似方法相比，他们内建了处理缺失值的功能。

In [37]: df = pd.DataFrame([[1.4, np.nan], [7.1, -4.5],
    ...:                    [np.nan, np.nan], [0.75, -1.3]],
    ...:                    index=['a', 'b', 'c', 'd'],
    ...:                    columns=['one', 'two'])

In [38]: df
Out[38]:
    one  two
a  1.40  NaN
b  7.10 -4.5
c   NaN  NaN
d  0.75 -1.3

sum 返回列和，axis='columns'或axis=1，返回行和

In [39]: df.sum()
Out[39]:
one    9.25
two   -5.80
dtype: float64

In [40]: df.sum(axis='columns')
Out[40]:
a    1.40
b    2.60
c    0.00
d   -0.55
dtype: float64

In [41]: df.sum(axis=1)
Out[41]:
a    1.40
b    2.60
c    0.00
d   -0.55
dtype: float64

通过禁用 skipna 实现不排除 NA 值

In [42]: df.mean(axis='columns', skipna=False)
Out[42]:
a      NaN
b    1.300
c      NaN
d   -0.275
dtype: float64

归约方法可选参数

方法	描述
axis	归约轴，0为行香，1为列项
skipna	排除缺失值，默认为 True
level	如果轴是多层索引的（MultiIndex），该参数可以缩减分组层级

idxmin 和 idxmax，返回间接统计信息，如最大值和最小值的索引值

In [1]: import numpy as np

In [2]: import pandas as pd

In [3]: df = pd.DataFrame([[1.4, np.nan], [7.1, -4.5],
   ...:                   [np.nan, np.nan], [0.75, -1.3]],
   ...:                   index=['a', 'b', 'c', 'd'],
   ...:                   columns=['one', 'two'])

In [4]: df
Out[4]:
    one  two
a  1.40  NaN
b  7.10 -4.5
c   NaN  NaN
d  0.75 -1.3

In [6]: df.idxmax()
Out[6]:
one    b
two    d
dtype: object

积累型方法

In [7]: df.cumsum()
Out[7]:
    one  two
a  1.40  NaN
b  8.50 -4.5
c   NaN  NaN
d  9.25 -5.8

describe 一次性产生多个汇总统计

In [8]: df.describe()
Out[8]:
            one       two
count  3.000000  2.000000
mean   3.083333 -2.900000
std    3.493685  2.262742
min    0.750000 -4.500000
25%    1.075000 -3.700000
50%    1.400000 -2.900000
75%    4.250000 -2.100000
max    7.100000 -1.300000

In [9]: obj = pd.Series(['a', 'a', 'b', 'c'] * 4)

In [10]: obj.describe()
Out[10]:
count     16
unique     3
top        a
freq       8

汇总统计及其相关方法的完整列表在《利用Python进行数据分析》159页