timedelta object to float

Question

Hello any ideas how can convert timedelta object to float in pandas it needs to generalize based on minutes or hours i.e 00:06:46.221000 to 6.46 minutes or 01:06:76.341000 to 66.76 minutes

thanks

Same idea as the duplicate but with minutes, divide by a minute instead of an hour — user3483203, Sep 22 '19 at 18:13

tsb5555 · Accepted Answer · 2019-09-22T17:30:04.863

1

If you are using Python2.7/3.2 or higher, you can use the following method (where time_delta_obj is your timedelta object):

time_minutes = time_delta_obj.total_seconds()/60.0

In Python 3.2 or higher, you could simply use this:

time_minutes = time_delta_obj/datetime.timedelta(minutes=1)

See Python 3.2 - datetime and time and timedelta.total_seconds() for more information.

To apply this to a pandas dataframe, use apply():

def convert_to_minutes(time_delta_obj):
    return time_delta_obj.total_seconds()/60.0

or

def convert_to_minutes(time_delta_obj):
    return time_delta_obj/datetime.timedelta(minutes=1)

Then for the column named 'time_diff' of the df use this:

df['time_diff'] = df['time_diff'].apply(convert_to_minutes)

edited Sep 22 '19 at 17:30

answered Sep 22 '19 at 17:10

tsb5555

321
3
13

thanks how do you apply this to a pandas series timedelta obj – goerge Sep 22 '19 at 17:23
@goerge: you divide it by a `Timedelta(minutes=1)`. – Willem Van Onsem Sep 22 '19 at 17:26

score 1 · Answer 2 · answered Sep 22 '19 at 17:25

1

You can just divide your Series with a Timedelta(minutes=1). Indeed:

>>> df
              col
0 00:06:46.221000
1 00:03:12.456000
2 00:02:07.123456
>>> df['col'] / pd.Timedelta(minutes=1)
0    6.770350
1    3.207600
2    2.118724
Name: col, dtype: float64

answered Sep 22 '19 at 17:25

Willem Van Onsem

443,496
30
428
555

timedelta object to float

2 Answers2