pandas.TimedeltaIndex.asi8#

property TimedeltaIndex.asi8[source]#

Return Integer representation of the values.

For DatetimeIndex and TimedeltaIndex, the values are the number of time units (determined by the index resolution) since the epoch. For PeriodIndex, the values are ordinals.

Returns:

numpy.ndarray: An ndarray with int64 dtype.

See also

Index.values: Return an array representing the data in the Index, using native types (datetime64, timedelta64) rather than int64.
Index.to_numpy: Return a NumPy ndarray of the index values.

Examples

For DatetimeIndex with default microsecond resolution:

>>> idx = pd.DatetimeIndex(["2023-01-01", "2023-01-02"], dtype="datetime64[us]")
>>> idx.asi8
array([1672531200000000, 1672617600000000])

For TimedeltaIndex with millisecond resolution:

>>> idx = pd.TimedeltaIndex(["1 day", "2 days"], dtype="timedelta64[ms]")
>>> idx.asi8
array([ 86400000, 172800000])

For PeriodIndex:

>>> idx = pd.PeriodIndex(["2023-01", "2023-02", "2023-03"], freq="M")
>>> idx.asi8
array([636, 637, 638])