I know about round, ceil, floor functions.
df.index.round("30min")
This rounds to the nearest 30 minute interval. What I want is that each is rounded to 30 minutes.
In the case of .round 10:15 will be rounded to 10:30 and 10:45 to 11:00. I would want both rounded to 10:30.
The datetime Index:
DatetimeIndex(['2021-11-29 09:30:00', '2021-11-29 09:45:00',
'2021-11-29 10:00:00', '2021-11-29 10:15:00',
'2021-11-29 10:30:00', '2021-11-29 10:45:00',
'2021-11-29 11:00:00', '2021-11-29 11:15:00',
'2021-11-29 11:30:00', '2021-11-29 11:45:00',
...
'2021-12-27 13:30:00', '2021-12-27 13:45:00',
'2021-12-27 14:00:00', '2021-12-27 14:15:00',
'2021-12-27 14:30:00', '2021-12-27 14:45:00',
'2021-12-27 15:00:00', '2021-12-27 15:15:00',
'2021-12-27 15:30:00', '2021-12-27 15:45:00'],
dtype='datetime64[ns]', name='Datetime', length=520, freq=None)
CodePudding user response:
Use DatetimeIndex.floor
by hour and add 30 minutes:
df = pd.DataFrame(index=pd.DatetimeIndex(
['2021-11-29 09:30:00', '2021-11-29 09:45:00',
'2021-11-29 10:00:00', '2021-11-29 10:15:00',
'2021-11-29 10:30:00', '2021-11-29 10:45:00',
'2021-11-29 11:00:00', '2021-11-29 11:15:00',
'2021-11-29 11:30:00', '2021-11-29 11:45:00',
'2021-12-27 13:30:00', '2021-12-27 13:45:00',
'2021-12-27 14:00:00', '2021-12-27 14:15:00',
'2021-12-27 14:30:00', '2021-12-27 14:45:00',
'2021-12-27 15:00:00', '2021-12-27 15:15:00',
'2021-12-27 15:30:00', '2021-12-27 15:45:00']))
print (df.index.floor("1H") pd.Timedelta('30min'))
DatetimeIndex(['2021-11-29 09:30:00', '2021-11-29 09:30:00',
'2021-11-29 10:30:00', '2021-11-29 10:30:00',
'2021-11-29 10:30:00', '2021-11-29 10:30:00',
'2021-11-29 11:30:00', '2021-11-29 11:30:00',
'2021-11-29 11:30:00', '2021-11-29 11:30:00',
'2021-12-27 13:30:00', '2021-12-27 13:30:00',
'2021-12-27 14:30:00', '2021-12-27 14:30:00',
'2021-12-27 14:30:00', '2021-12-27 14:30:00',
'2021-12-27 15:30:00', '2021-12-27 15:30:00',
'2021-12-27 15:30:00', '2021-12-27 15:30:00'],
dtype='datetime64[ns]', freq=None)