如何提取单个字符串并删除其他类似于DataFrame的字符串

Question

我将类别名称与技能名称结合起来，按类别名称对其进行排序。现在我有了列如下所示的表

(Category1) Skill 1
(Category1) Skill 2
(Category1) Skill 3
(Category1) Skill 4
(Category1) Skill 5
(Category1) Skill 6
(Category2) Skill 7
(Category2) Skill 8
(Category2) Skill 9
(Category2) Skill 10
(Category2) Skill 11
(Category2) Skill 12

我只想为每个技能保留一个类别标题，并删除另一个，类似于有类似于此表的表。

(Category1) Skill 1
Skill 2
Skill 3
Skill 4
Skill 5
Skill 6
(Category2) Skill 7
Skill 8
Skill 9
Skill 10
Skill 11
Skill 12

有什么想法吗？谢谢

Answer 1

您可以拆分字符串并检索最后一部分Skill x，以及检查Categoryx被复制的位置，并使用结果替换为拆分的部分：

import numpy as np

m = df.col1.str.split(r'\) ', expand=True)
df['col1'] = np.where(m.duplicated(subset=0), m[1], df.col1)

               col1
0   (Category1) Skill 1
1               Skill 2
2               Skill 3
3               Skill 4
4               Skill 5
5               Skill 6
6   (Category2) Skill 7
7               Skill 8
8               Skill 9
9              Skill 10
10             Skill 11
11             Skill 12

输入数据-

 col1
0    (Category1) Skill 1
1    (Category1) Skill 2
2    (Category1) Skill 3
3    (Category1) Skill 4
4    (Category1) Skill 5
5    (Category1) Skill 6
6    (Category2) Skill 7
7    (Category2) Skill 8
8    (Category2) Skill 9
9   (Category2) Skill 10
10  (Category2) Skill 11
11  (Category2) Skill 12

Answer 2

假设您的dataframe(df)列名为'A'：

df2 = df.A.str.split(expand=True)
df2[0]=df2[0].mask(df2[0].eq(df2[0].shift())).fillna('')]
df.A = df2.apply(lambda x: ' '.join(x), axis=1)