Accumulate specific rows from pandas dataframe

Question

My Pandas dataframe looks like this:

Notice that for each task there are exactly 2 rows: row for begin and and row for end of task.
I need to create another dataframe with columns Task, StartTime, EndTime where each row will represent task and it's lifetime.
There is an ugly way to do it (using iteration). Is there an elegant way?

sort by action -> groupby task ->apply/cast to a pd.Series -> unstack — Dillon Davis, May 10 '22 at 02:42

score 1 · Accepted Answer · answered May 10 '22 at 03:30

1

Using pandas.pivot() function you can simply do it like this assuming df is your dataframe:

new_df = df.pivot(index="Task", columns="Action", values="DateTime")

answered May 10 '22 at 03:30

Tawfiq Abuaita

59
4

score 1 · Answer 2 · answered May 10 '22 at 04:24

Another way is to use set_index and unstack:

>>> (df.set_index(['Task', 'Action'])['DateTime'].unstack()
       .reset_index().rename_axis(columns=None))

   Task                    begin                      end
0     1  2022-05-04 16:07:42.580  2022-05-04 16:07:53.565

Accumulate specific rows from pandas dataframe

2 Answers2