[FIXED] Selecting first row of each index in multi indexing pandas DataFrame

Issue

choose every first row of each index of multiindexing pandas dataframe.

grouped = ecommerce[["category_id", "brand", "price"]].groupby(by=["category_id", "brand"]).mean()
grouped_sort = grouped.sort_values(by=["category_id", "price"], ascending=False)

grouped_sort looks like this:

now on this data frame I want to choose in each category just the first brands with the highest price.

Can anyone help me?

Solution

The following code can help:

gsgb = grouped_sort.copy()
gsgb = gsgb.groupby(level=0)
print(type(gsgb))
gsgb.head()

for cat, df in gsgb:
    display(df.sort_values(by=["price"], ascending=False).reset_index().iloc[0])

Working:
It basically loops over all categories in the grouped dataframe and then sort values based on price, then reset index and finally choose the one with max price.

Answered By - sotmot