Pandas groupby timestamp and increase count - python-3.x

Question

I am a beginner in pandas and I would like some help about a problem I have.
I have a csv file structured as follow:
#timestamp. message. name. ID
2021-07-10 14:01:00 user 0001 has logged out. User Log Off. 0001
2021-07-10 14:01:10 user 0002 has logged out. User Log Off. 0002
2021-07-10 14:01:15 user 0003 has logged out. User Log Off. 0003
2021-07-10 14:08:20 user 0001 has logged out. User Log Off. 0001
I would like to do, is to go through all the columns, and check if they are doubles, and if they are double in a time span of 10 min(based on the timestamp) to add a column with the number of the counted event.
for example this is what I would like to have as an output
#timestamp. message. name. ID. count
2021-07-10 14:01:00 user 0001 has logged out. User Log Off. 0001. 2
2021-07-10 14:01:10 user 0002 has logged out. User Log Off. 0002. 1
2021-07-10 14:01:15 user 0003 has logged out. User Log Off. 0003 1
Basically group the double event into only one row with the number of event counted in that time span.
Is this something achievable with pandas?
Thank you so much for any help

Answer 1

Here's an outline that you can follow:
# 0. sort data by timestamp if not already sorted
df = df.sort_values('#timestamp')
# lazy groupby
groups = df.groupby(['message.','name.', 'ID'])
# 1. compute the time differences `timediff` and compare to threshold
df['timediff'] = groups['#timstamp.'].diff() > pd.Timedelta('10T')
# 2. find the blocks with cumsum
df['block'] = groups['timediff'].cumsum()
# 3. groupby the blocks
out = (df.groupby(['blocks','message.','name.', 'ID'])
.agg({'#timestamp.':'first', 'timediff':'count'})
)
Note this will group 00:00:00, 00:09:00, and 00:18:00 together.