Pandas - drop_duplicates with multiple conditions

Question

I have a dataset where I want to remove duplicates based on some conditions.

For example, say I have a table as

ID  date    group
3001    2010    DCM
3001    2012    NII
3001    2012    DCM

I wanna say look into ID column for the similar IDs, if two dates were similar keep the row that group is NII

so it would become

ID  date    group
3001    2010    DCM
3001    2012    NII

Answer 1

Leverage duplicated here:

df[~df.duplicated(['ID', 'date'], keep=False) | df['group'].eq('NII')]

     ID  date group
0  3001  2010   DCM
1  3001  2012   NII