CLN: Remove Series._from_array #19893

jaumebonet · 2018-02-25T09:40:50Z

This PR transfers the class check of Series._from_array to DataFrame._box_col_values; removing the need for calling the Series._from_array and allowing for user-defined slice-inheritance of metadata.

I opted to delete the function as it was private but leave the Series.from_array public access working and it's FutureWarning of deprecation, as it was. To do that, I had to modify Series.from_array so it would keep also its original functionality until is completely deleted.

closes Deprecate Series._from_array ? #19883
passed tests.
passes git diff upstream/master -u -- "*.py" | flake8 --diff

Transfer its check to DataFrame._box_col_values

until is properly deleted

codecov · 2018-02-25T10:56:58Z

Codecov Report

Merging #19893 into master will increase coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master   #19893      +/-   ##
==========================================
+ Coverage   91.65%   91.66%   +<.01%     
==========================================
  Files         150      150              
  Lines       48915    48939      +24     
==========================================
+ Hits        44835    44859      +24     
  Misses       4080     4080

Flag	Coverage Δ
#multiple	`90.04% <100%> (ø)`	⬆️
#single	`41.81% <90%> (-0.02%)`	⬇️

Impacted Files	Coverage Δ
pandas/core/sparse/series.py	`95.2% <ø> (-0.05%)`	⬇️
pandas/core/dtypes/concat.py	`99.16% <100%> (+0.01%)`	⬆️
pandas/core/series.py	`94.2% <100%> (-0.21%)`	⬇️
pandas/core/frame.py	`97.18% <100%> (-0.05%)`	⬇️
pandas/core/dtypes/cast.py	`87.68% <0%> (-0.3%)`	⬇️
pandas/core/generic.py	`95.88% <0%> (-0.05%)`	⬇️
pandas/core/indexes/datetimelike.py	`97.38% <0%> (+0.01%)`	⬆️
pandas/core/indexes/timedeltas.py	`90.99% <0%> (+0.05%)`	⬆️
pandas/core/groupby.py	`92.31% <0%> (+0.1%)`	⬆️
... and 2 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 10cc8f4...e4ec987. Read the comment docs.

pep8speaks · 2018-02-25T11:09:37Z

Hello @jaumebonet! Thanks for updating the PR.

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

Comment last updated on February 26, 2018 at 14:55 Hours UTC

jreback · 2018-02-25T15:38:50Z

pandas/tests/frame/test_subclass.py

        assert not isinstance(result, tm.SubclassedDataFrame)
        tm.assert_series_equal(result, expected)
+
+    def test_frame_subclassing_and_inherit(self):


what is the purpose of these tests? this is another issue, yes?

this should be in a new PR

I wrote it to check that that approach would work now. I'll take it out.

jreback · 2018-02-25T15:47:17Z

pandas/core/frame.py

+        # This check here was previously performed in Series._from_array
+        # By doing it here there is no need for that function anymore
+        # GH#19883.
+        from pandas.core.dtypes.generic import ABCSparseArray


make this a function like _get_series_result_type and put it near _get_frame_result_type in pandas/core/dtypes/concat.py, it should just take self and return the type.

I see the idea.
Wouldn't it make more sense to be in pandas/core/dtypes/commons.py ? As that file is already imported to frame.py?

no, common is for introspection, not conversion

jreback · 2018-02-25T15:47:53Z

pandas/core/frame.py

        """ provide boxed values for a column """
-        return self._constructor_sliced._from_array(values, index=self.index,
-                                                    name=items, fastpath=True)
+        # This check here was previously performed in Series._from_array


remove all of this commentary

to check type of DataFrame._constructor_sliced

jreback · 2018-02-25T21:03:37Z

pandas/core/dtypes/concat.py



+def _get_sliced_frame_result_type(data, obj):
+    """


add a full doc-string

jreback · 2018-02-25T21:03:46Z

pandas/core/dtypes/concat.py

+    if is_sparse(data):
+        from pandas.core.sparse.api import SparseSeries
+        return SparseSeries
+    else:


just return, no else needed

jreback · 2018-02-25T21:04:05Z

pandas/core/dtypes/concat.py

+    the data is sparse or not.
+    """
+    if is_sparse(data):
+        from pandas.core.sparse.api import SparseSeries


from pandas import

jreback · 2018-02-25T21:04:26Z

pandas/core/frame.py

        """ provide boxed values for a column """
-        return self._constructor_sliced._from_array(values, index=self.index,
-                                                    name=items, fastpath=True)
+        from pandas.core.dtypes.concat import _get_sliced_frame_result_type


this can be imported at the top

jreback · 2018-02-25T21:04:39Z

pandas/core/frame.py

-        return self._constructor_sliced._from_array(values, index=self.index,
-                                                    name=items, fastpath=True)
+        from pandas.core.dtypes.concat import _get_sliced_frame_result_type
+        this_constructor_sliced = _get_sliced_frame_result_type(values, self)


call this klass

jreback

small changes, ping on green

jreback · 2018-02-26T13:26:45Z

pandas/core/dtypes/concat.py

+
+    Parameters
+    ----------
+    data : ndarray


jreback · 2018-02-26T13:26:54Z

pandas/core/dtypes/concat.py

+    Series or SparseSeries
+    """
+    if is_sparse(data):
+        from pandas.core.sparse.api import SparseSeries


from pandas import SparseSeries

jaumebonet · 2018-02-26T13:41:23Z

done!

jreback · 2018-02-26T13:45:26Z

thanks ping on green.

jaumebonet · 2018-02-26T14:01:32Z

sorry... I'm not sure what "ping on green" means in this context 😅

jreback · 2018-02-26T14:15:19Z

it means watch this page and report when all of the checks are green.

jreback · 2018-02-27T01:15:36Z

thanks @jaumebonet

jaumebonet added 3 commits February 25, 2018 10:24

Remove Series._from_array

ae676b8

Transfer its check to DataFrame._box_col_values

keep original function of Series.from_array

0c67e5a

until is properly deleted

variable name error fix

3d6a4dd

add test_frame_subclassing_and_inherit

e27554c

pep8 fixes

c643922

jreback requested changes Feb 25, 2018

View reviewed changes

jreback added Sparse Sparse Data Type Clean labels Feb 25, 2018

_get_sliced_frame_result_type

545de1e

to check type of DataFrame._constructor_sliced

jreback requested changes Feb 25, 2018

View reviewed changes

style fixes

f000fdb

jreback requested changes Feb 26, 2018

View reviewed changes

style updates

356b02b

jreback added this to the 0.23.0 milestone Feb 26, 2018

jreback approved these changes Feb 26, 2018

View reviewed changes

rollback import in _get_sliced_frame_result_type

e4ec987

jreback merged commit ceb9031 into pandas-dev:master Feb 27, 2018

harisbal pushed a commit to harisbal/pandas that referenced this pull request Feb 28, 2018

CLN: Remove Series._from_array (pandas-dev#19893)

69f0e8b

Uh oh!

CLN: Remove Series._from_array #19893

CLN: Remove Series._from_array #19893

Uh oh!

Conversation

jaumebonet commented Feb 25, 2018 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

codecov bot commented Feb 25, 2018 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

pep8speaks commented Feb 25, 2018 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Comment last updated on February 26, 2018 at 14:55 Hours UTC

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jreback left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jaumebonet commented Feb 26, 2018

Uh oh!

jreback commented Feb 26, 2018

Uh oh!

jaumebonet commented Feb 26, 2018

Uh oh!

jreback commented Feb 26, 2018

Uh oh!

jreback commented Feb 27, 2018

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

jaumebonet commented Feb 25, 2018 •

edited

Loading

codecov bot commented Feb 25, 2018 •

edited

Loading

pep8speaks commented Feb 25, 2018 •

edited

Loading