Simplify to_pydatetime()

[jbrockmendel]

Up until #17331, Timestamp.microsecond was very slow. So previously it was faster to go through convert_to_tsobject to get a new datetime instance than it was to just return
datetime(self.year, self.month, self.day, self.hour, self.minute, self.second, self.microsecond, self.tzinfo)

Now it is slightly faster (and much simpler) to do it this way.

Before:

In [2]: ts = pd.Timestamp.now()
In [3]: %timeit ts.to_pydatetime()
The slowest run took 13.88 times longer than the fastest. This could mean that an intermediate result is being cached.
1000000 loops, best of 3: 1.44 µs per loop

After:

In [2]: ts = pd.Timestamp.now()
In [3]: %timeit ts.to_pydatetime()
The slowest run took 8.51 times longer than the fastest. This could mean that an intermediate result is being cached.
1000000 loops, best of 3: 1.29 µs per loop

With a tz-aware Timestamp the speedup is much bigger:

Before:

In [4]: import pytz
In [5]: ts2 = ts.replace(tzinfo=pytz.timezone('US/Pacific'))
In [6]: %timeit ts2.to_pydatetime()
The slowest run took 39.71 times longer than the fastest. This could mean that an intermediate result is being cached.
10000 loops, best of 3: 68.9 µs per loop

After:

In [7]: import pytz
In [8]: ts2 = ts.replace(tzinfo=pytz.timezone('US/Pacific'))
In [9]: %timeit ts2.to_pydatetime()
The slowest run took 8.29 times longer than the fastest. This could mean that an intermediate result is being cached.
1000000 loops, best of 3: 1.47 µs per loop

[codecov]

Codecov Report

Merging #17592 into master will decrease coverage by 0.01%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master   #17592      +/-   ##
==========================================
- Coverage   91.19%   91.18%   -0.02%     
==========================================
  Files         163      163              
  Lines       49627    49627              
==========================================
- Hits        45259    45250       -9     
- Misses       4368     4377       +9

Flag	Coverage Δ
#multiple	88.96% <ø> (ø)	⬆️
#single	40.19% <ø> (-0.07%)	⬇️

Impacted Files	Coverage Δ
pandas/io/gbq.py	25% <0%> (-58.34%)	⬇️
pandas/core/frame.py	97.77% <0%> (-0.1%)	⬇️

---

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 b3087ef...235c8d0. Read the comment docs.

[chris-b1]

indendation

[chris-b1]

LGTM, minus formatting nit.

[jreback]

can you verify that there is an asv for this (and add one if not).

[jbrockmendel]

There is not, just added one.

[pep8speaks]

Hello @jbrockmendel! Thanks for updating the PR.

Cheers ! There are no PEP8 issues in this Pull Request. 🍻

Comment last updated on September 21, 2017 at 22:48 Hours UTC

[jbrockmendel]

Looks like someone added one overnight, will add to_pydatetime timing to it.

[jbrockmendel]

Is there an error message in the circleci log that I'm missing?

[jreback]

looks fine

you can click thru to circleci to see errors

[jbrockmendel]

looks fine

you can click thru to circleci to see errors

I don't ask just because I want to waste your time. I'm saying I don't see any error message in the log.

[jreback]

I don't ask just because I want to waste your time. I'm saying I don't see any error message in the log.

I know, but some people don't actually realize you can click thru. yeah it prob timed out or something.

Wait till I merge #17619 and rebase

[jreback]

thanks!