Pandas To datetime & Cumsum Milliseconds

Question

I have a several question.

First, I want to datetime in pandas dataframe.

like this... 2018/03/06 00:01:27:744

How can I replace this datetime?

And Then.. Second..

        Time                  Sensor1         Sensor2    TimeCumsum
2018/03/06 00:01:27:744         0               1            
2018/03/06 00:01:27:759         0               1
2018/03/06 00:01:27:806         0               1            0.15
2018/03/06 00:01:27:838         1               1    
2018/03/06 00:01:28:009         1               1            0.2
2018/03/06 00:01:28:056         1               0            ...

I wanna Time Seconds cumsum when Sensor1 is 0 And Sensor2 is 1.

How can I do this?

Thanks.

[The docs](https://pandas.pydata.org/pandas-docs/stable/generated/pandas.to_datetime.html) for time and [the docs](https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.resample.html) for resample to help you get a cumsum. You'll want to do something like ```df = df.assign(seconds = df.dt.seconds)``` to get a column of seconds and then use ```.resample``` or ```.cumsum``` to get your answer. — lwileczek, May 08 '18 at 12:21
Is that to_datetime use this?? milliseconds is that correct?? :( .. — GrayHash, May 08 '18 at 12:27

score 1 · Answer 1 · answered May 08 '18 at 12:28

1

I believe need:

df['Time'] = pd.to_datetime(df['Time'], format='%Y/%m/%d %H:%M:%S:%f')

m = (df['Sensor1'].eq(0) & df['Sensor2'].eq(1))
df['col'] = df.loc[m, 'Time'].dt.microsecond.cumsum() // 10**3
print (df)
                     Time  Sensor1  Sensor2     col
0 2018-03-06 00:01:27.744        0        1   744.0
1 2018-03-06 00:01:27.759        0        1  1503.0
2 2018-03-06 00:01:27.806        0        1  2309.0
3 2018-03-06 00:01:27.838        1        1     NaN
4 2018-03-06 00:01:28.009        1        1     NaN
5 2018-03-06 00:01:28.056        1        0     NaN

answered May 08 '18 at 12:28

jezrael

822,522
95
1,334
1,252

Assume... But I see this Error... "ValueError: cannot reindex from a duplicate axis".. Don't add to df.. Can you help me again? Thans jezrael. – GrayHash May 08 '18 at 12:36
@GrayHash - What is `print (df.index)` ? – jezrael May 08 '18 at 12:41
@GrayHash - Can you explain `0.15` in expected output? – jezrael May 08 '18 at 12:42
Sorry, Not Correct df at my dataframe. length is not same. How can I give to df that means... em.. df[df["Sensor1]==0 & df["Sensor2"]==1] Is that right? – GrayHash May 08 '18 at 12:44
Maye I dont understand `cumsum`, `0.15` and `0.2` value. Can you explain more? – jezrael May 08 '18 at 12:45
so.. sensor1 is 0 value & sensor 2 is 1 value cumsum that times. maybe I think that 10**7 not 10**3. – GrayHash May 08 '18 at 12:48
Let us [continue this discussion in chat](https://chat.stackoverflow.com/rooms/170617/discussion-between-jezrael-and-grayhash). – jezrael May 08 '18 at 12:49
How Can I column "col" Initialize that 0?? – GrayHash May 16 '18 at 12:30
Do you think `df['col'] -= df['col'].min()` ? – jezrael May 16 '18 at 12:31
No. your Answer is cumsum all col column. But I want to Initialize col column when Sensor value is 1. – GrayHash May 16 '18 at 12:34
@GrayHash - i dont understand, can you add excected output? – jezrael May 16 '18 at 12:35
Can you See that 6, 7 rows? – GrayHash May 16 '18 at 12:38
Thank you, why first value `744` is not `0` ? – jezrael May 16 '18 at 12:40
That's Output Data. Original Data is not haven't col column. – GrayHash May 16 '18 at 12:47
Let us [continue this discussion in chat](https://chat.stackoverflow.com/rooms/171158/discussion-between-jezrael-and-grayhash). – jezrael May 16 '18 at 12:48

Pandas To datetime & Cumsum Milliseconds

1 Answers1