FutureWarning：使用“datetime64[ns]”dtype 将时区感知的 DatetimeArray 转换为时区天真的 ndarray

Question

I upgraded from pandas 0.20.3 to pandas 0.24.1. While running the command ts.sort_index(inplace=True) , I am getting a FutureWarning in my test output, which is shown below. Can I change the method call to suppress the following warning? I am happy to keep the old behavior.

/lib/python3.6/site-packages/pandas/core/sorting.py:257: FutureWarning: Converting timezone-aware DatetimeArray to timezone-naive ndarray with 'datetime64[ns]' dtype. In the future, this will return an ndarray with 'object' dtype where each element is a 'pandas.Timestamp' with the correct 'tz'.
    To accept the future behavior, pass 'dtype=object'.
    To keep the old behavior, pass 'dtype="datetime64[ns]"'.
  items = np.asanyarray(items)

My index looks like the following prior to running the sort_index:

ts.index
DatetimeIndex(['2017-07-05 07:00:00+00:00', '2017-07-05 07:15:00+00:00',
               '2017-07-05 07:30:00+00:00', '2017-07-05 07:45:00+00:00',
               ...
               '2017-07-05 08:00:00+00:00'],
              dtype='datetime64[ns, UTC]', name='start', freq=None)

Answer 1

I rewrote your question here , to include an MCVE. After it went a while with no responses, I posted an issue against Pandas.

Here's my workaround:

with warnings.catch_warnings():
    # Bug in Pandas emits useless warning when sorting tz-aware index
    warnings.simplefilter("ignore")
    ds = df.sort_index()

Answer 2

If I were you, I would do a downgrade using pip and setting the previous version. It's the lazier answer. But if you really want to keep it upgraded, then there is a parameter call deprecated warning inside pandas data frame. Just adjust it accordingly what you need. You can check it using the documentation of pandas. Have a nice night