ENH: PeriodIndex now has period dtype #13941
Conversation
sinhrks
added
Dtype Conversions
|
Looks superb! Thanks @sinhrks |
sinhrks
force-pushed
the
period_dtype
branch
2 times, most recently
from
f9740ed to
af2def9
Compare
| ``PeriodIndex`` now has ``period`` dtype | ||
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ | ||
|
|
||
| ``PeriodIndex`` now has its own ``period`` dtype. ``period`` dtype is a |
Current coverage is 85.30% (diff: 92.80%)@@ master #13941 diff @@
==========================================
Files 139 139
Lines 50241 50346 +105
Methods 0 0
Messages 0 0
Branches 0 0
==========================================
+ Hits 42851 42947 +96
- Misses 7390 7399 +9
Partials 0 0
|
|
What do you get when you write |
|
@shoyer yeah that should return an object array of Periods (similar to what a datetime tz does) |
removed unused / duplicated internal method returning array-likes. Also added some tests to guarantee existing API before fixing its dtype (see #13941). Author: sinhrks <[email protected]> Closes #13955 from sinhrks/object_array_cln and squashes the following commits: a75a718 [sinhrks] CLN: Period cleanup related to array like meth
|
split |
sinhrks
force-pushed
the
period_dtype
branch
3 times, most recently
from
537b7e4 to
e2b4564
Compare
|
@sinhrks a couple of minor changes, but otherwise lgtm. period is currently broken as its |
| In [1]: pi = pd.PeriodIndex(['2016-08-01'], freq='D') | ||
| In [2]: pi | ||
| Out[2]: PeriodIndex(['2016-08-01'], dtype='int64', freq='D') | ||
| In [3]: pd.types.api.is_integer_dtype(pi) |
There was a problem hiding this comment.
This is not the current public entry point (although depends on outcome of discussion in #13634), I think pd.api.types.is_.. is
There was a problem hiding this comment.
yep.. @jorisvandenbossche right here
| dtype similar to the timezone aware dtype (``datetime64[ns, tz]``). | ||
|
|
||
| The ``period`` dtype holds the ``freq`` attribute and is represented with | ||
| ``period[freq]``, using :ref:`frequency strings <timeseries.offset_aliases>`. |
There was a problem hiding this comment.
add e.g. period[D] or period[M]
|
just some minor changes / perf checks. Getting the |
| .. _timeseries.timezone_series: | ||
|
|
||
| The ``period`` dtype holds the ``freq`` attribute and is represented with | ||
| ``period[freq]`` like ``period[D]`` or ``period[D]``, using :ref:`frequency strings <timeseries.offset_aliases>`. |
There was a problem hiding this comment.
Second period[D] should be a different example?
|
thanks @sinhrks nice change! |
|
Thx for review. Because #14010 includes some period related tests, it'll broken on master. I'll send a fix soon. |
xref #13941. Author: sinhrks <[email protected]> Closes #14025 from sinhrks/test_coercion_period and squashes the following commits: b873683 [sinhrks] TST: Fix test_coercion for period dtype
split from #13941 (comment) Author: sinhrks <[email protected]> Closes #13988 from sinhrks/period_values and squashes the following commits: d7637c9 [sinhrks] API: PeriodIndex.values now return array of Period objects
git diff upstream/master | flake8 --diffsplitted
dtyperelated part from #13755, to clarify the differences related toPeriodIndex