BUG: DataFrame.quantile fails on datetime values · Issue #6965 · pandas-dev/pandas (original) (raw)

In [2]: df = DataFrame({'a': pd.to_datetime(['2010', '2011']), 'b': [0, 5]})

In [3]: df.quantile(.5, numeric_only=False) Out[3]: a 1.278072e+18 b 1.278072e+18 dtype: float64

Series handles things fine.

We need to do a view('i8') on the date time array somewhere, but I had trouble just sticking it inside of the f that gets passed to data.apply(f, ...)