Sum the values of a column based on the timestamp

Asked Feb 05 '19 at 16:12

Active Feb 05 '19 at 16:15

Viewed 116 times

My dataframe looks like this:

           timestamp    side    price   amount  fee
0   2018-12-14 13:00:00 buy  0.00338508 5   0.005
1   2018-12-14 13:00:00 buy  0.00337829 5   0.005
2   2018-12-14 14:00:00 buy  0.00338089 5   0.005
3   2018-12-14 14:00:00 buy  0.00338079 1   0.001
4   2018-12-14 15:00:00 sell 0.00338838 8   0.00002711
5   2018-12-14 15:00:00 sell 0.00339132 8   0.00002713
6   2018-12-14 15:00:00 buy  0.00337999 5   0.005
7   2018-12-14 16:00:00 sell 0.00339373 5   0.00001697

What I need to do is to create a column which would sum the values of the column 'amount' which have the same timestamp. So for the timestamp 2018-12-14 13:00:00 the new column would look like this:

          timestamp    side    price   amount  fee sum
0   2018-12-14 13:00:00 buy  0.00338508 5   0.005 10
1   2018-12-14 13:00:00 buy  0.00337829 5   0.005 10

I tried with the following code: df_statistics_hour['traded_volume'] = float(df_statistics_hour['amount'].sum()) but of course there is the same value in each cell of the new column, any idea on how to proceed? Thanks!

edited Feb 05 '19 at 16:15

glibdud

7,550
4
27
37

asked Feb 05 '19 at 16:12

Viktor.w

1,787
2
20
46

Use `df_statistics_hour['traded_volume'] = df_statistics_hour.groupby('timestamp')['amount'].transform('sum')` – jezrael Feb 05 '19 at 16:14
1

Yup indeed, thanks! – Viktor.w Feb 05 '19 at 17:03

Sum the values of a column based on the timestamp

0 Answers0