使用另一列

时间:2017-12-08 05:10:17

标签: python pandas

我有一个pandas数据框:

+--------+---------+------+----------------+
|  Name  | Address |  ID  |   Linked_To    |
+--------+---------+------+----------------+
| Name A | ABC     | 1233 | 1234;1235      |
| Name B | DEF     | 1234 | 1233;1236;1237 |
| Name C | GHI     | 1235 | 1234;1233;2589 |
+--------+---------+------+----------------+

Linked_To列中的某些ID是Name列下的记录。我可以创建一个字典并将Linked_To列中的数据作为列表传递。但是,我不确定如何继续。理想情况下,我希望看到类似的内容:

+--------+---------+------+-------------------------+
|  Name  | Address |  Id  |        Linked To        |
+--------+---------+------+-------------------------+
| Name A | ABC     | 1233 | Name B;Name C           |
| Name B | DEF     | 1234 | Name A;Name D; Name E   |
| Name C | HIJ     | 1235 | Name B;Name A; None     |
+--------+---------+------+-------------------------+

1 个答案:

答案 0 :(得分:1)

没有一些循环似乎很难做到这一点:

linked = df.Linked_To.str.split(';')

def pull_name(iden):
    try:
        return df[df.ID == int(iden)].Name.iat[0]
    except:
        return str(None)

res = linked.apply(lambda ids: '; '.join([pull_name(i) for i in ids]))

print(res)
0         Name B; Name C
1     Name A; None; None
2    Name B; Name A; ...
Name: Linked_To, dtype: object