根据条件选择列中的最大值

Question

I have two columns. ID and Percentage. Some IDs are not unique. Assume I have IDs 233, 233, 277, 277. And corresponding percentages: 4,5%, 7%, 3%, 1%. I need to select max. percentage for each ID. So that outcome is: 233 - 7%, 277 - 3%.

I wrote code that returns max value for the whole column, not the specific non-unique ID.

df['help_column'] = np.where(df.duplicated() ==True, max(df['percentage']),0)

As the highest value in the whole column is 33%, I get 33% for ID 233, and 33% for ID 277 instead of desired result. Thanks

Answer 1

这更像是一种transform

df['help_column'] = df.groupby('ID')['percentage'].transform('max')

Answer 2

尝试这个

df.groupby(['ID'])['percentage'].max()