Grouby with datetime in two dimensions

Question

I am checking this thread already, but somehow it is not working.

I currently have a dataframe with two columns:

When I check the column types I get:

Datum        datetime64[ns]
MIPS                float64

So now I want to groupby the day and the hour:

df.groupby([df['Datum'].dt.day, df['Datum'].dt.hour])['MIPS'].sum().reset_index()

But I get the following error:

ValueError: cannot insert Datum, already exists

Is it because he cannot create a grouped dataframe with two times the same name? How can I avoid that?

Answer 1

You can rename both Series for avoid MultiIndex with same names:

df.groupby([df['Datum'].dt.day.rename('Day'), 
            df['Datum'].dt.hour.rename('Hour')])['MIPS'].sum().reset_index()