Python Pandas groupby月日年周

Question

I have a pandas DataFrame that looks like this:

Day    Month Year       Date Week Data1 Data2
  1  January 2016 01-01-2016    1    15    22
  2  January 2016 01-02-2016    1    16    26
 12  January 2016 01-12-2016    2    18    29
  8 February 2016 02-08-2016    2     9    17

I manually added the week column and they are type int. I want to group the data by week, month, year ie: January 2016 week 1, January 2016 week 2, February 2016 week 2. The issue is I want the result to be a DataFrame where Data1 is the sum of the columns and Data2 is the mean of the columns.

Thank you for the help.

Answer 1

I think you can use groupby followed by agg for your desired aggregations.

df.groupby(['Year', 'Month', 'Week']).agg({'Data1':'sum', 'Data2':'mean'})

Demo

>>> df.groupby(['Year', 'Month', 'Week']).agg({'Data1':'sum', 'Data2':'mean'})

                    Data1  Data2
Year Month    Week              
2016 February 2         9     17
     January  1        31     24
              2        18     29