如何根据另一列对一行中的特定列求和？

Question

could you please help my with the following issue:

Sample File :

I try to aggregate the sum of 6 months, up from a specific start date per row.

The sum should be shown in a new column (Sum 6 months from startdate)

My first thought would be to get it with the following code:

df['sum_6_months'] = df.loc[:,'01.2018':'06.2018'].apply(sum, axis=1)

but this code is not individually and only for the timeframe (01.18-06.18) in all rows.

df = pd.DataFrame(np.array([[1, 5, 3, 4, 5, 6, 7, 7, 8,2,5,7,3,4,1], [1, 5, 3, 4, 5, 6, 7, 7,8,2,5,7,3,4,2],[1,5,3, 3, 4, 5, 6, 7, 7, 8,2,5,7,3,4],
                             [1, 5, 3, 4, 5, 6, 7, 7, 8,2,5,7,3,4,3], [1, 5, 3, 4, 5, 6, 7, 7,8,2,5,7,3,4,4],[1,5,5, 3, 4, 5, 6, 7, 7, 8,2,5,7,3,4],
                             [1, 5, 3, 4, 5, 6, 7, 7, 8,2,5,7,3,4,5], [1, 5, 3, 4, 5, 6, 7, 7,8,2,5,7,3,4,5],[1,5,2, 3, 4, 5, 6, 7, 7, 8,2,5,7,3,4],
                             [1, 5, 3, 4, 5, 6, 7, 7, 8,2,5,7,3,4,6], [1, 5, 3, 4, 5, 6, 7, 7,8,2,5,7,3,4,2],[1,5,5, 3, 4, 5, 6, 7, 7, 8,2,5,7,3,4],
                             [1, 5, 3, 4, 5, 6, 7, 7, 8,2,5,7,3,4,4], [1, 5, 3, 4, 5, 6, 7, 7,8,2,5,7,3,4,2],[1,5,1, 3, 4, 5, 6, 7, 7, 8,2,5,7,3,4]]),
                   columns=['01.2018', '02.2018', '03.2018', '04.2018', '05.2018','06.2018', '07.2018', '08.2018',
                            '09.2018','10.2018', '11.2018', '12.2018','01.2019', '02.2019', '03.2019'])

date = [01.2018, 03.2018,04.2018,05.2018,03.2018,01.2018, 03.2018,04.2018,05.2018,03.2018,01.2018, 03.2018,04.2018,05.2018,03.2018]
df['Startdate']= date

Answer 1

First, calculate the number of columns to skip in each row:

df2['StartCol'] = 1 + df2.columns[1:].searchsorted(df2.Startdate)

The 1: skips the Startdate column. Then "roll" that many columns to the left, so they wrap around and end up at the end of each row, take the first 6, and sum:

np.roll(df2.iloc[:, 1:], -df2.StartCol)[:,:6].sum(1)

That gives you:

[27, 28, 27, 21, 26, 27, 29, 27, 21, 25, 27, 25, 25, 18, 23]

Which you can store in a new column if you like.

Answer 2

df['Startdate']=df['Startdate'].astype(str).str.rjust(7,'0')

df_columns = df.columns.tolist()
def get_sum_six(df_list):
    start_date_index = df_columns.index(df_list[-1])
    df_list = df_list[0:-1]
    sum_of_six = sum(df_list[start_date_index: start_date_index + min(len(df_list)-start_date_index, 6)])
    return (sum_of_six)

df['sum_last_six'] = df.apply(lambda x: get_sum_six(x.tolist()), axis=1)