pandas.DatetimeIndex.snap#

DatetimeIndex.snap(freq='S')[source]#

Snap time stamps to nearest occurring frequency.

Deprecated since version 3.2.0: For fixed frequencies snap returns the values unchanged; use DatetimeIndex.round() if rounding was intended. For other offsets, use::

offset = pd.tseries.frequencies.to_offset(freq) back, fwd = idx.map(offset.rollback), idx.map(offset.rollforward) fwd.where(fwd - idx <= idx - back, back)

Each timestamp in the index is adjusted to the nearest occurrence of the specified frequency, snapping backward or forward as appropriate.

Parameters:

freqstr, Timedelta, datetime.timedelta, or DateOffset, default ‘S’: Frequency strings can have multiples, e.g. ‘5h’. See here for a list of frequency aliases.

Returns:

DatetimeIndex: Time stamps to nearest occurring freq.

See also

DatetimeIndex.round: Perform round operation on the data to the specified freq.
DatetimeIndex.floor: Perform floor operation on the data to the specified freq.

Examples

>>> idx = pd.DatetimeIndex(
...     ["2023-01-01", "2023-01-02", "2023-02-01", "2023-02-02"],
...     dtype="M8[ns]",
... )
>>> idx
DatetimeIndex(['2023-01-01', '2023-01-02', '2023-02-01', '2023-02-02'],
dtype='datetime64[ns]', freq=None)
>>> idx.snap("MS")
DatetimeIndex(['2023-01-01', '2023-01-01', '2023-02-01', '2023-02-01'],
dtype='datetime64[ns]', freq=None)