如何计算在列值上传递的行的时间差

Question

I have a pandas dataframe like

Status	Time Stamp
Passing	2021-11-25 15:15:36
Failing	2021-11-25 00:46:23
Failing	2021-11-25 00:16:03
Failing	2021-11-24 23:45:08
Passing	2021-11-25 15:15:13
Failing	2021-11-25 00:46:47
Failing	2021-11-25 00:16:09
Failing	2021-11-24 23:44:59

I need to get the time of the first passing event to the first instance of when it failed for that sequence. So the difference row 0 and row 3 , and add it to a new column.

Then I need it to calculate the next sequence and add it to the value in the new column.

So the difference between row 4 and row 7 and add the difference to the previous time so I get the total time it was failing.

This is what the df should look like at the end

Status	Time Stamp	Downtime	Total Downtime
Passing	2021-11-25 15:15:36	15:30:38	31:00:52
Failing	2021-11-25 00:46:23	15:30:38	31:00:52
Failing	2021-11-25 00:16:03	15:30:38	31:00:52
Failing	2021-11-24 23:45:08	15:30:38	31:00:52
Passing	2021-11-25 15:15:13	15:30:14	31:00:52
Failing	2021-11-25 00:46:47	15:30:14	31:00:52
Failing	2021-11-25 00:16:09	15:30:14	31:00:52
Failing	2021-11-24 23:44:59	15:30:14	31:00:52

Note that this is example data and the index's of passing and failing events are at different index each time.

Here is my code

import pandas as pd

data = {'Status': ['Passing','Failing','Failing','Failing','Passing','Failing','Failing','Failing'],

'TimeStamp': ['2021-11-25 15:15:36','2021-11-25 00:46:23','2021-11-25 00:16:03','2021-11-24 23:45:08','2021-11-25 15:15:13','2021-11-25 00:46:47','2021-11-25 00:16:09','2021-11-24 23:44:59']}

df = pd.DataFrame(data)

I'm self taught in Python and pandas and have no idea how to achieve what I need. Any help would be appreciated.

Answer 1

You did not specify at which index do you mean to add the resulting difference of passing and failing. the oldestFailing below include all the differences that you need.

from datetime import datetime as dt
prevPassIdx = 0
prevOldest = dt.now()
oldestFailing = []
for i in range(1, len(df['TimeStamp'])):
    if df['Status'][i] == 'Passing':
        prevPassIdx = i
        oldestFailing.append(prevOldest)
        prevOldest = dt.now()
    else:
        if dt.strptime(df['TimeStamp'][i],"%Y-%m-%d %H:%M:%S") <prevOldest:
            prevOldest = dt.strptime(df['TimeStamp'][i],"%Y-%m-%d %H:%M:%S")
if df['Status'][i] != ('Passing'):
    oldestFailing.append(prevOldest) 

Here, oldestFailing output is as follows:

[datetime.datetime(2021, 11, 24, 23, 45, 8),
 datetime.datetime(2021, 11, 24, 23, 44, 59)]