大熊猫保持重复的最高价值

Question

I have data similar to:

id value duplicate
a   200  yes
a   12   yes
b   42   yes
c   12   no
b   532  yes
b   21   yes
...

To track the duplicates I use df['duplicate'] = df.duplicated('id', keep=False) However, I would like to keep the ones with the highest value and either mark or drop the other duplicates. Any suggestions?

Answer 1

Ah I don't know why I didn't think of this first. df.sort(['id', 'value']) df['is_duplicated'] = df.duplicated('id', keep='first')

sorry!