提取子字符串并转换为datetime python

Question

我有一个数据帧，df：

data = [{0: 18, 1: '(Responses) 17th Nov 20'},
        {0: 304, 1: '(Responses) 17th Nov 20'},
        {0: 1177, 1: '(Responses) 17th Nov 20'},
        {0: 899, 1: '(Responses) 17th Nov 20'}]

df = pd.DataFrame(data)

0                1                                          
18    (Responses) 17th Nov 20
304   (Responses) 17th Nov 20
1177  (Responses) 17th Nov 20
899   (Responses) 17th Nov 20

有没有有效的方法提取出17th Nov 2020并将其转换为新的column[2]as17-11-2020as日期

对于其他日期，它也可以是1st或2nd或3rd

预期产出：

0                1                 2                         
18    (Responses) 17th Nov 20   17-11-2020
304   (Responses) 17th Nov 20   17-11-2020
1177  (Responses) 17th Nov 20   17-11-2020
899   (Responses) 17th Nov 20   17-11-2020

Answer 1

只需split以“（responses）”作为关键字的字符串，然后获取拆分后的第二个元素：

  df['new_column'] = df['1'].str.split("(responses)").str[1]

Answer 2

尝试使用str.split和pd.to_datetime：

df[2] = pd.to_datetime(df[1].str.replace('\(Responses\) ', ''))
print(df)

输出：

      0                        1          2
0    18  (Responses) 17th Nov 20 2020-11-17
1   304  (Responses) 17th Nov 20 2020-11-17
2  1177  (Responses) 17th Nov 20 2020-11-17
3   899  (Responses) 17th Nov 20 2020-11-17