list first N rows from Pandas df using groupby-CodePudding

I am using the following code to group a pandas df, and pass the values grouped to a list

  df.groupby(['Column1','Column2'])['Column3'].apply(list)

This returns all values from column3 in the list.
How can i get the list to include only first N examples only from column3 ?

CodePudding user response：

Use indexing with lambda function:

df.groupby(['Column1','Column2'])['Column3'].apply(lambda x: list(x)[:N])

Or:

df.groupby(['Column1','Column2'])['Column3'].apply(lambda x: list(x.iloc[:N]))

EDIT:

df.groupby(['Column1','Column2'])['Column3'].apply(lambda x: list(x.unique()[:N]))

CodePudding user response：

You could use str accessor to get the first N elements from lists:

df.groupby(['Column1','Column2'])['Column3'].agg(list).str[:N]