pandas.DatetimeIndex.set_freq

DatetimeIndex.set_freq(freq)

Return a new Index with the given frequency.

The values are unchanged; only the freq attribute differs. Unlike assigning to freq, this does not modify the original Index.

Parameters:

freqstr, DateOffset or None

Frequency to set. It must conform to the values of the Index. "infer" sets the frequency inferred from the values, or None if none can be inferred. None removes the frequency.

Returns:

DatetimeIndex or TimedeltaIndex

A new Index with freq set.

Raises:

ValueError

If freq is invalid or does not conform to the values of the Index.

TypeError

If the Index is an empty TimedeltaIndex and freq is not a fixed frequency.

See also

DatetimeIndex.freq

Return the frequency object if it is set.

DatetimeIndex.inferred_freq

Return the frequency inferred from the values.

Examples

>>> idx = pd.DatetimeIndex(["2020-01-01", "2020-01-03", "2020-01-05"])
>>> idx
DatetimeIndex(['2020-01-01', '2020-01-03', '2020-01-05'],
              dtype='datetime64[us]', freq=None)
>>> idx.set_freq("2D")
DatetimeIndex(['2020-01-01', '2020-01-03', '2020-01-05'],
              dtype='datetime64[us]', freq='2D')
>>> idx.set_freq("2D").set_freq(None)
DatetimeIndex(['2020-01-01', '2020-01-03', '2020-01-05'],
              dtype='datetime64[us]', freq=None)
>>> idx.set_freq("infer")
DatetimeIndex(['2020-01-01', '2020-01-03', '2020-01-05'],
              dtype='datetime64[us]', freq='2D')