DOC/DEPR: pivot_annual

[sinhrks]

• closes Create a generic version of convert_to_annual function in scikits.timeseries #736
• tests added / passed
• passes git diff upstream/master | flake8 --diff
• whatsnew entry

Updated cookbook attaching SO link of original issue.

BTW, pivot_annual has a similar functionality but it looks to be replaced by pivot_table and datetime property access. Because the function is not on top pd namespace, shouldn't deprecate it?

[jreback]

yes this lgtm. @jorisvandenbossche any objections?

[codecov-io]

Current coverage is 84.54%

Merging #13706 into master will decrease coverage by <.01%

@@             master     #13706   diff @@
==========================================
  Files           141        141          
  Lines         51185      51164    -21   
  Methods           0          0          
  Messages          0          0          
  Branches          0          0          
==========================================
- Hits          43275      43255    -20   
+ Misses         7910       7909     -1   
  Partials          0          0          

Powered by Codecov. Last updated by 506520b...d097bab

[jorisvandenbossche]

No real objections from me, although the functions (looking at the length of the code) seems to do a lot more than the cookbook example?
SO gives no results for a search for pivot_annual, a good sign :-)

Deprecate isleapyear as well?

[jreback]

I agree should deprecate isleapyear but add on as an accessor for Timestamp/DTI and .dt
which should be straightforward to do

[jorisvandenbossche]

@jreback There were still some unresolved comments I think (deprecating isleapyear as well)

[jreback]

@jorisvandenbossche ahh yes, I mean to make an issue about that.

[sinhrks]

I wii subbmit new PR (to deprecate isleapyear and adding the same proprty).

one minor pount is the property name. Timestamp has property like .is_year_end, numpy datetime.c has is_leapyear on the other hand. Which is prefereble, is_leap_year or is_leapyear?

[jorisvandenbossche]

no problem!

[jreback]

I think is_leapyear is fine

leapyear is really the common english word (I see though already getting inconsistent as we should really do is_year_leap)?