Below you can find multiple ways to round to the nearest minute or seconds in Pandas:

(1) rounding pandas timestamp

pd.to_datetime(df['date']).dt.round(freq='30min').dt.time

result:

0    11:00:00
1    12:30:00
2    12:00:00
3    15:30:00
Name: date, dtype: object

(2) round date to 30 minutes python

df['date'].dt.round('15min')

results:

0   2024-09-01 11:00:00
1   2024-09-01 12:30:00
2   2024-09-01 12:15:00
3   2024-09-01 15:45:00
Name: date, dtype: datetime64[ns]

(3) rounding pandas timestamps to the nearest minute or 5 minutes

df['date'].dt.round('5min')

0   2024-09-01 11:05:00
1   2024-09-01 12:35:00
2   2024-09-01 12:15:00
3   2024-09-01 16:00:00
Name: date, dtype: datetime64[ns]

(4) round to a minute in python/pandas

df['date'].dt.round("10s")

result:

0   2024-09-01 11:04:40
1   2024-09-01 12:35:20
2   2024-09-01 12:16:30
3   2024-09-01 15:58:00
Name: date, dtype: datetime64[ns]

(4) floor to a closest minute

df['date'].dt.floor('30min')

result:

0   2024-09-01 11:00:00
1   2024-09-01 12:30:00
2   2024-09-01 12:00:00
3   2024-09-01 15:30:00
Name: date, dtype: datetime64[ns]

(5) ceil to a closest minute

df['date'].dt.ceil('30min')

result:

0   2024-09-01 11:30:00
1   2024-09-01 13:00:00
2   2024-09-01 12:30:00
3   2024-09-01 16:00:00
Name: date, dtype: datetime64[ns]

Different frequencies to round

  • d - day
  • h - hour
  • min - minute
  • s - second

Sample Data

import pandas as pd

data = {
    'date': [
        pd.Timestamp('2024-09-01 11:04:45'),
        pd.Timestamp('2024-09-01 12:035:15'),
        pd.Timestamp('2024-09-01 12:16:30'),
        pd.Timestamp('2024-09-01 15:57:59')
    ]
}

df = pd.DataFrame(data)
df

Resource