PERF: EWMA with times

[mroeschke]

• closes PERF: ewm passed with times is so slow compared to ewm pased with int #39784
• tests added / passed
• Ensure all linting tests pass, see here for how to run them
• whatsnew entry

In [1]: idx=pd.date_range('20000101','20201231',periods=50000)
   ...:
   ...: df=pd.DataFrame(data=range(50000),index=idx)

In [2]: %timeit df.ewm(halflife=pd.Timedelta('100d'),times=df.index).mean()

1.2 ms ± 13.8 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each) <-- PR

10.2 s ± 32.2 ms per loop (mean ± std. dev. of 7 runs, 1 loop each) <-- master

[jreback]

very nice!

[jorisvandenbossche]

This seems to have resulted in a performance regression for the existing ASV benchmark: https://pandas.pydata.org/speed/pandas/#rolling.EWMMethods.time_ewm?python=3.8&Cython=0.29.21&p-constructor='DataFrame'&p-constructor='Series'&p-window=1000&p-dtype='float'&p-method='mean'&commits=ab687aec-f066cda6

[simonjayhawkins]

opened #42333 to keep track of this regression.