Pandas: get Item ID of the most popular in a category (two Groupby level)

Asked Jan 28 '21 at 06:22

Active Jan 28 '21 at 06:32

Viewed 40 times

I'm sorry that I cannot think of a title.

I have a DF like this:

I want to get the item that is the most popular among the item category item_cat1. This is the desired output:

Explanation: in item_cat1 A, item_id 3 is sold the most (4) compares to item_id 1 (2).

I tried train.groupby(["item_cat1", "item_id"])["item_count"].sum(), but I don't know how to choose only the max value.

P/S: what I want is the item_id, not the item_count. This answer does not help me: Get the row(s) which have the max value in groups using groupby

edited Jan 28 '21 at 06:27

asked Jan 28 '21 at 06:22

Minh-Long Luu

2

Dupe answer `train.loc[train.groupby(["item_cat1"])["item_count"].idxmax()]` not working? – jezrael Jan 28 '21 at 06:25
2

If not working, can you add more data for see how `train.loc[train.groupby(["item_cat1"])["item_count"].idxmax()]` failed? Now in sample data working very well – jezrael Jan 28 '21 at 06:29
Thank you, it works! It is my bad that the DF can contain NaNs. – Minh-Long Luu Jan 28 '21 at 06:40
Super! Happy coding! – jezrael Jan 28 '21 at 06:41

0 Answers0