我正在尝试合并多个数据框，以找到他们应该没有的用户组。
DF1
AccountType Name SID disabled compname localaccount lockout passwordchangeable passwprdexpires passwordrequired status
512 Administrator 1001 True mycoolguycompname True True True NONE True Degraded
512 coolguy 500 False mycoolguycompname True False True NONE True OK
512 Visitor 501 True mycoolguycompname True False False NONE True Degraded
DF2
groupname localaccount GroupSID sidtype status members
Administrators True S-1-5-32-544 4 OK [coolguy, cool_Admins, Inspector_Cool, cool_2, Administrator, Backup Operators]
Backup Operators True S-1-5-32-551 4 OK [coolguy, badguy]
Certificate Service DCOM Access True S-1-5-32-574 4 OK [cool_2]
Cryptographic Operators True S-1-5-32-569 4 OK [Administrators, Visitor]
我如何合并两个数据框以显示与它们所属的适当组正确相关的用户名，如下所示：
预期产量：
usernames groups GroupSID status compname
coolguy Administrators S-1-5-32-544 OK mycoolguycompname
cool_Admins Administrators S-1-5-32-544 OK mycoolguycompname
Inspector_Cool Administrators S-1-5-32-544 OK mycoolguycompname
cool_2 Administrators S-1-5-32-544 OK mycoolguycompname
Administrator Administrators S-1-5-32-544 OK mycoolguycompname
coolguy Backup Operators S-1-5-32-551 OK mycoolguycompname
badguy Backup Operators S-1-5-32-551 OK mycoolguycompname
cool_2 Certificate Service DCOM Access S-1-5-32-574 OK mycoolguycompname
coolguy Cryptographic Operators S-1-5-32-569 OK mycoolguycompname
cool_Admins Cryptographic Operators S-1-5-32-569 OK mycoolguycompname
Inspector_Cool Cryptographic Operators S-1-5-32-569 OK mycoolguycompname
cool_2 Cryptographic Operators S-1-5-32-569 OK mycoolguycompname
Administrator Cryptographic Operators S-1-5-32-569 OK mycoolguycompname
Visitor Cryptographic Operators S-1-5-32-569 OK mycoolguycompname
badguy Cryptographic Operators S-1-5-32-569 OK mycoolguycompname
我非常担心嵌套的组块。任何帮助将不胜感激。
最佳答案
IIUC您可以通过以下方式进行操作：
In [49]: d1 = df1[['Name','compname']]
...: d2 = df2[['groupname','GroupSID','members','status']]
...:
...: lst_col = 'members'
...:
...: pd.DataFrame({
...: col:np.repeat(d2[col].values, d2[lst_col].str.len())
...: for col in d2.columns.difference([lst_col])
...: }).assign(**{'Name':np.concatenate(d2[lst_col])}).merge(d1, on='Name', how='left')
...:
...:
Out[49]:
GroupSID groupname status Name compname
0 S-1-5-32-544 Administrators OK coolguy mycoolguycompname
1 S-1-5-32-544 Administrators OK cool_Admins NaN
2 S-1-5-32-544 Administrators OK Inspector_Cool NaN
3 S-1-5-32-544 Administrators OK cool_2 NaN
4 S-1-5-32-544 Administrators OK Administrator mycoolguycompname
5 S-1-5-32-544 Administrators OK Backup Operators NaN
6 S-1-5-32-551 Backup Operators OK coolguy mycoolguycompname
7 S-1-5-32-551 Backup Operators OK badguy NaN
8 S-1-5-32-574 Certificate Service DCOM Access OK cool_2 NaN
9 S-1-5-32-569 Cryptographic Operators OK Administrators NaN
10 S-1-5-32-569 Cryptographic Operators OK Visitor mycoolguycompname