[ENH] Use default EA repr for IntervalArray #26316

makbigc · 2019-05-08T11:55:19Z

closes Use default EA repr for IntervalArray #25022
1 test deleted
passes git diff upstream/master -u -- "*.py" | flake8 --diff
whatsnew entry

codecov · 2019-05-08T12:29:35Z

Codecov Report

Merging #26316 into master will decrease coverage by 51.3%.
The diff coverage is n/a.

@@             Coverage Diff             @@
##           master   #26316       +/-   ##
===========================================
- Coverage   92.04%   40.74%   -51.31%     
===========================================
  Files         175      175               
  Lines       52291    52291               
===========================================
- Hits        48132    21306    -26826     
- Misses       4159    30985    +26826

Flag	Coverage Δ
#multiple	`?`
#single	`40.74% <ø> (-0.15%)`	⬇️

Impacted Files	Coverage Δ
pandas/core/arrays/interval.py	`38.15% <ø> (-54.92%)`	⬇️
pandas/core/indexes/interval.py	`35.33% <ø> (-59.9%)`	⬇️
pandas/io/formats/latex.py	`0% <0%> (-100%)`	⬇️
pandas/io/sas/sas_constants.py	`0% <0%> (-100%)`	⬇️
pandas/core/groupby/categorical.py	`0% <0%> (-100%)`	⬇️
pandas/tseries/plotting.py	`0% <0%> (-100%)`	⬇️
pandas/tseries/converter.py	`0% <0%> (-100%)`	⬇️
pandas/io/formats/html.py	`0% <0%> (-99.37%)`	⬇️
pandas/io/sas/sas7bdat.py	`0% <0%> (-91.16%)`	⬇️
pandas/io/sas/sas_xport.py	`0% <0%> (-90.1%)`	⬇️
... and 132 more

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 23b6115...95c012a. Read the comment docs.

codecov · 2019-05-08T12:29:36Z

Codecov Report

❗ No coverage uploaded for pull request base (master@be4b48e). Click here to learn what that means.
The diff coverage is 100%.

@@            Coverage Diff            @@
##             master   #26316   +/-   ##
=========================================
  Coverage          ?   91.97%           
=========================================
  Files             ?      180           
  Lines             ?    50774           
  Branches          ?        0           
=========================================
  Hits              ?    46701           
  Misses            ?     4073           
  Partials          ?        0

Flag	Coverage Δ
#multiple	`90.62% <100%> (?)`
#single	`41.83% <100%> (?)`

Impacted Files	Coverage Δ
pandas/core/arrays/interval.py	`92.19% <100%> (ø)`
pandas/core/indexes/interval.py	`95.95% <100%> (ø)`

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 be4b48e...645e01e. Read the comment docs.

TomAugspurger · 2019-05-08T15:23:03Z

I'm not sure we need / want to make this change.

jschendel · 2019-05-10T22:42:55Z

Also not sure if we need/want this change, as it adds an additional layer of complexity to maintaining common docstrings. Certainly something we could work around if we want to change the repr but not ideal.

cc @jorisvandenbossche : what are your thoughts here?

jschendel

There are a some additional IntervalArray docstrings that would need to be updated to match the new repr that aren't currently being checked by CI, e.g. from_arrays, from_breaks, etc.

pandas/pandas/core/arrays/interval.py

Lines 252 to 257 in bb0376c

    
               Examples 
        
               -------- 
        
               >>> pd.%(qualname)s.from_breaks([0, 1, 2, 3]) 
        
               %(klass)s([(0, 1], (1, 2], (2, 3]], 
        
                             closed='right', 
        
                             dtype='interval[int64]')

jschendel · 2019-05-10T22:45:02Z

pandas/core/indexes/interval.py

@@ -1086,8 +1086,52 @@ def equals(self, other):
                self.right.equals(other.right) and
                self.closed == other.closed)

-    @Appender(_interval_shared_docs['overlaps'] % _index_doc_kwargs)


I think there should be a way of reusing the main portion of the docstring, and appending in custom examples. Something similar to the constructor maybe?

pandas/pandas/core/indexes/interval.py

Lines 100 to 124 in bb0376c

@Appender(_interval_shared_docs['class'] % dict(

klass="IntervalIndex",

summary="Immutable index of intervals that are closed on the same side.",

name=_index_doc_kwargs['name'],

versionadded="0.20.0",

extra_attributes="is_overlapping\nvalues\n",

extra_methods="contains\n",

examples=textwrap.dedent("""\

Examples

--------

A new ``IntervalIndex`` is typically constructed using

:func:`interval_range`:

>>> pd.interval_range(start=0, end=5)

IntervalIndex([(0, 1], (1, 2], (2, 3], (3, 4], (4, 5]],

closed='right',

dtype='interval[int64]')

It may also be constructed using one of the constructor

methods: :meth:`IntervalIndex.from_arrays`,

:meth:`IntervalIndex.from_breaks`, and :meth:`IntervalIndex.from_tuples`.

See further examples in the doc strings of ``interval_range`` and the

mentioned constructor methods.

"""),

jorisvandenbossche · 2019-05-11T12:44:41Z

Personally I think in principle it would be nice to start using the same repr style for all Array classes.
But the docstring sharing is indeed something we first need to think through.
How do we currently do that for DatetimeArray? We don't share docstrings with DatetimeIndex?

jreback · 2019-05-11T19:16:32Z

@TomAugspurger what is your objection here? this de-special cases IntervalArray repr and makes it like all other Arrays, a good thing.

TomAugspurger · 2019-05-11T19:34:08Z

Having to rewrite the docstrings to pass doctests.

…

On May 11, 2019, at 14:16, Jeff Reback ***@***.***> wrote: @TomAugspurger what is your objection here? this de-special cases IntervalArray repr and makes it like all other Arrays, a good thing. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

jreback · 2019-06-08T20:30:54Z

@makbigc can you merge master.

makbigc · 2019-06-12T12:15:42Z

working. Give me some more time.

makbigc · 2019-07-19T12:35:18Z

@jschendel Amended. Please tell me if anything is wrong.

jreback · 2019-07-21T00:28:04Z

ok this looks pretty good now; @jschendel

jschendel

Can you add a whatsnew note with previous/current behavior examples to the 1.0 whatsnew since this is technically a backwards incompatible change?

jschendel · 2019-07-24T19:39:02Z

Will review this in more depth later tonight but looks good overall. Would be nice to have an automated way of converting between IntervalArray and IntervalIndex example sections but that can be saved for a follow up.

WillAyd · 2019-08-28T16:47:57Z

doc/source/whatsnew/v1.0.0.rst

+
+.. ipython:: python
+
+   In [2]: pd.arrays.IntervalArray.from_tuples([(0, 1), (2, 3)])


Since this is an actual python block I don't think you need the In/Out - that may be causing the CI failure

@WillAyd
Thanks for your advice. Otherwise I can't pass the CI tests.

pandas/tests/arrays/interval/test_interval.py

makbigc · 2019-09-04T16:34:28Z

@WillAyd @jschendel Anything else?

WillAyd · 2019-09-05T17:21:02Z

doc/source/whatsnew/v1.0.0.rst

+
+*pandas 1.0.0*
+
+.. code-block:: ipython


Sorry could have been clearer but just make this `.. ipython:: python`` for the directive. You shouldn't then need the In / Out below and it will render automatically for you (this is typically what we do with new behavior)

WillAyd · 2019-09-06T15:40:30Z

doc/source/whatsnew/v1.0.0.rst

+
+*pandas 1.0.0*
+
+.. code-block:: python


In case it helps you should use code-block for the old behavior and the ipython directive for newly introduced behavior. You can see an example of this with the Better Repr for MultiIndex section back in the 0.25 whatsnew

https://github.com/pandas-dev/pandas/blob/5f31bca3813b2ed71363a9d40518f217d2b546d6/doc/source/whatsnew/v0.25.0.rst#better-repr-for-multiindex

WillAyd

lgtm @jschendel @TomAugspurger @jorisvandenbossche

jschendel · 2019-09-10T02:29:01Z

thanks @makbigc!

makbigc added 4 commits May 8, 2019 19:48

Use new repr and fix docstring

c2ab939

Delete one test

f9740c7

Remove relundant code

54bc56b

Remove relundant import

95c012a

gfyoung added ExtensionArray Extending pandas with custom dtypes or arrays. Interval Interval data type labels May 8, 2019

jschendel suggested changes May 10, 2019

View reviewed changes

makbigc added 2 commits June 20, 2019 15:26

merge

e7b865c

Separate the example section from _interval_shared_docs

cc25182

makbigc force-pushed the ENH-25022 branch 2 times, most recently from 23322a6 to c7121a7 Compare June 22, 2019 09:48

Fix docstring error

4c86f53

makbigc force-pushed the ENH-25022 branch from c7121a7 to 4c86f53 Compare June 22, 2019 10:13

makbigc added 3 commits July 14, 2019 22:41

resolve conflict

474b4c4

Fix the example in contain method

0e07474

Merge for update

c9dea18

makbigc force-pushed the ENH-25022 branch from 645e01e to c9dea18 Compare July 19, 2019 08:49

makbigc added 2 commits July 19, 2019 17:44

Fix the black format

55f701a

Fix docstring error

480d780

jreback added this to the 1.0 milestone Jul 21, 2019

jschendel suggested changes Jul 24, 2019

View reviewed changes

makbigc added 4 commits August 18, 2019 22:47

merge for update

3bc839e

Add whatsnew entry in v1.0

3c36719

Fix lint error

0a62694

Fix black error

f6d4f03

WillAyd requested changes Aug 28, 2019

View reviewed changes

makbigc added 4 commits August 30, 2019 12:31

Fix doc format error and add test_repr

32ddfa9

4th merge for update

0db48c3

Fix format error

5a16fa7

Remove tailing space in v1.0 rst

64889ec

WillAyd requested changes Sep 5, 2019

View reviewed changes

makbigc added 2 commits September 6, 2019 21:49

Fix v1.0.0.rst

5ab6ae9

remove the output in v1.0

5f31bca

WillAyd reviewed Sep 6, 2019

View reviewed changes

Fix v1

6bbee48

WillAyd approved these changes Sep 6, 2019

View reviewed changes

jschendel approved these changes Sep 10, 2019

View reviewed changes

jschendel merged commit 96bf661 into pandas-dev:master Sep 10, 2019

makbigc deleted the ENH-25022 branch September 10, 2019 04:14

proost pushed a commit to proost/pandas that referenced this pull request Dec 19, 2019

[ENH] Use default EA repr for IntervalArray (pandas-dev#26316)

081320d

proost pushed a commit to proost/pandas that referenced this pull request Dec 19, 2019

[ENH] Use default EA repr for IntervalArray (pandas-dev#26316)

14783cc

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[ENH] Use default EA repr for IntervalArray #26316

[ENH] Use default EA repr for IntervalArray #26316

makbigc commented May 8, 2019 •

edited

codecov bot commented May 8, 2019

codecov bot commented May 8, 2019 •

edited

TomAugspurger commented May 8, 2019

jschendel commented May 10, 2019 •

edited

jschendel left a comment

jschendel May 10, 2019

jorisvandenbossche commented May 11, 2019

jreback commented May 11, 2019

TomAugspurger commented May 11, 2019 via email

jreback commented Jun 8, 2019

makbigc commented Jun 12, 2019

makbigc commented Jul 19, 2019

jreback commented Jul 21, 2019

jschendel left a comment

jschendel commented Jul 24, 2019

WillAyd Aug 28, 2019

makbigc Aug 30, 2019

makbigc commented Sep 4, 2019

WillAyd Sep 5, 2019

WillAyd Sep 6, 2019

WillAyd left a comment

jschendel commented Sep 10, 2019

	Examples
	--------
	>>> pd.%(qualname)s.from_breaks([0, 1, 2, 3])
	%(klass)s([(0, 1], (1, 2], (2, 3]],
	closed='right',
	dtype='interval[int64]')

	@Appender(_interval_shared_docs['class'] % dict(
	klass="IntervalIndex",
	summary="Immutable index of intervals that are closed on the same side.",
	name=_index_doc_kwargs['name'],
	versionadded="0.20.0",
	extra_attributes="is_overlapping\nvalues\n",
	extra_methods="contains\n",
	examples=textwrap.dedent("""\
	Examples
	--------
	A new ``IntervalIndex`` is typically constructed using
	:func:`interval_range`:

	>>> pd.interval_range(start=0, end=5)
	IntervalIndex([(0, 1], (1, 2], (2, 3], (3, 4], (4, 5]],
	closed='right',
	dtype='interval[int64]')

	It may also be constructed using one of the constructor
	methods: :meth:`IntervalIndex.from_arrays`,
	:meth:`IntervalIndex.from_breaks`, and :meth:`IntervalIndex.from_tuples`.

	See further examples in the doc strings of ``interval_range`` and the
	mentioned constructor methods.
	"""),


		.. ipython:: python

		In [2]: pd.arrays.IntervalArray.from_tuples([(0, 1), (2, 3)])

[ENH] Use default EA repr for IntervalArray #26316

[ENH] Use default EA repr for IntervalArray #26316

Conversation

makbigc commented May 8, 2019 • edited

codecov bot commented May 8, 2019

Codecov Report

codecov bot commented May 8, 2019 • edited

Codecov Report

TomAugspurger commented May 8, 2019

jschendel commented May 10, 2019 • edited

jschendel left a comment

Choose a reason for hiding this comment

jschendel May 10, 2019

Choose a reason for hiding this comment

jorisvandenbossche commented May 11, 2019

jreback commented May 11, 2019

TomAugspurger commented May 11, 2019 via email

jreback commented Jun 8, 2019

makbigc commented Jun 12, 2019

makbigc commented Jul 19, 2019

jreback commented Jul 21, 2019

jschendel left a comment

Choose a reason for hiding this comment

jschendel commented Jul 24, 2019

WillAyd Aug 28, 2019

Choose a reason for hiding this comment

makbigc Aug 30, 2019

Choose a reason for hiding this comment

makbigc commented Sep 4, 2019

WillAyd Sep 5, 2019

Choose a reason for hiding this comment

WillAyd Sep 6, 2019

Choose a reason for hiding this comment

WillAyd left a comment

Choose a reason for hiding this comment

jschendel commented Sep 10, 2019

makbigc commented May 8, 2019 •

edited

codecov bot commented May 8, 2019 •

edited

jschendel commented May 10, 2019 •

edited