问题

In [1]: import re
   ...: import pandas as pd
   ...: data = {
   ...:     'Dave': 
   ...:     'Steve': 
   ...:     'Rob': 
   ...:     'Wes': np.nan
   ...: }
   ...: data = pd.Series(data)
   ...: pattern = '([A-Z0-9._%+-]+)@([A-Z0-9.-]+)\\.([A-Z]{2,4})'
   ...: matches = data.str.match(pattern, flags=re.IGNORECASE)

In [2]: matches
Out[2]:
Dave     True
Steve    True
Rob      True
Wes       NaN
dtype: object

In [3]: matches.str.get(1)
Out[3]:
Dave    NaN
Steve   NaN
Rob     NaN
Wes     NaN
dtype: float64

In [4]: matches.str[0]
Out[4]:
Dave    NaN
Steve   NaN
Rob     NaN
Wes     NaN
dtype: float64

看到为什么matches已经匹配到数据，但是Out[3] 和 Out[4]获取数据的时候还是显示NaN呢？

我的解题思路

1. 首先，我做了一件事，先明白当前是什么对象，我们看到 matches 是一个 pandas.core.series.Series 对象

In [5]: type(matches)
Out[5]: pandas.core.series.Series

In [6]: s = pd.Series([
    ...:     "String",
    ...:     (1, 2, 3),
    ...:     ["a", "b", "c"],
    ...:     123,
    ...:     -456,
    ...:     { 1: "Hello", "2": "World" },
    ...:     True,
    ...:     False
    ...: ])

In [7]: type(s)     # 对比前面的In [5]，matches和现在的s是一样的类型
Out[7]: pandas.core.series.Series

In [8]: s.str.get(1)
Out[8]:
0        t  # "String",
1        2  # (1, 2, 3),
2        b  # ["a", "b", "c"],
3      NaN  # 123, 不可迭代
4      NaN  # -456, 不可迭代
5    Hello  # { 1: "Hello", "2": "World" },
6      NaN  # True, 不可迭代
7      NaN  # False, 不可迭代
dtype: object

其实并非所有对象 在 pandas.Series.str.get 方法调用下都能获取，必须是可迭代的对象才能获取

结论

pandas.core.series.Series对象 如果想调用 pandas.Series.str.get方法获取数据，Series的数据必须是可迭代的（字符串，元组，列表，字典）