COMPAT: NaT accessors

[jreback]

closes #15781

PR

In [1]: idx = DatetimeIndex(['1/1/2000', None, None, '1/4/2000'])

In [2]: idx.is_month_start
Out[2]: array([ True, False, False, False], dtype=bool)

In [3]: idx.is_leap_year
Out[3]: array([ True, False, False,  True], dtype=bool)

0.19.2

In [1]: idx = DatetimeIndex(['1/1/2000', None, None, '1/4/2000'])

In [3]: idx.is_month_start
Out[3]: array([  1.,  nan,  nan,   0.])

In [2]: idx.is_leap_year
Out[2]: array([ True, False, False,  True], dtype=bool)

mimics is_leap_year to return a boolean (inlcuding for NaT).

[jorisvandenbossche]

assert result is False
In the bare assert style for pytest, I find that a bit clearer if you are explicitly testing for False, and in this case, more importantly, otherwise 0 would pass the test (eg if for some reason the result if converted to float which is what happens in the current code)

[jreback]

yep fixed

[jreback]

@jorisvandenbossche

[codecov]

Codecov Report

Merging #15782 into master will decrease coverage by 0.01%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master   #15782      +/-   ##
==========================================
- Coverage   90.99%   90.98%   -0.02%     
==========================================
  Files         143      143              
  Lines       49386    49401      +15     
==========================================
+ Hits        44939    44946       +7     
- Misses       4447     4455       +8

Impacted Files	Coverage Δ
pandas/tseries/period.py	92.73% <100%> (+0.05%)	⬆️
pandas/tseries/common.py	88.09% <100%> (ø)	⬆️
pandas/tseries/tdi.py	90.33% <100%> (+0.09%)	⬆️
pandas/tseries/index.py	95.42% <100%> (+0.02%)	⬆️
pandas/io/gbq.py	25% <0%> (-58.34%)	⬇️
pandas/core/frame.py	97.56% <0%> (-0.1%)	⬇️
pandas/core/common.py	91.3% <0%> (+0.33%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 80280ec...6bc9bc0. Read the comment docs.

[jreback]

any comments?

[jorisvandenbossche]

Looks good!