BUG: conversion of Series to Categorical #16557
Conversation
|
@jreback this fixes the issue, could you review the PR and let me know if am doing it wrong. |
Codecov Report
@@ Coverage Diff @@
## master #16557 +/- ##
=======================================
Coverage 90.96% 90.96%
=======================================
Files 161 161
Lines 49263 49263
=======================================
Hits 44810 44810
Misses 4453 4453
Continue to review full report at Codecov.
|
|
pls add tests. in fact you want to write the tests first to make sure that they fail, then your fix makes it pass. also add a whatsnew entry (0.20.2) is fine. |
ca3404b to
3e1d73f
Compare
pandas/tests/series/test_dtypes.py
Outdated
| series = Series(['a', 'b', 'c']) | ||
|
|
||
| categorical_series = Series(series, dtype='category') | ||
| expected = Categorical(['a', 'b', 'c']) |
There was a problem hiding this comment.
I think you can mak this expected = Series(['a', 'b', 'c'], dtype='category')
and then you can use tm.assert_series_equal(categorical_series, expected).
27a955c to
83dd40a
Compare
|
@TomAugspurger I have made the changes according to your comments, the build is failing for some reason, can you retrigger the travis build please? |
|
could you retrigger circleci build as well please. Thank you. |
|
@jreback @TomAugspurger Could you review the PR please, Thank you. |
doc/source/whatsnew/v0.20.2.txt
Outdated
| @@ -59,6 +59,7 @@ Conversion | |||
|
|
|||
| - Bug in :func:`to_numeric` in which empty data inputs were causing a segfault of the interpreter (:issue:`16302`) | |||
| - Silence numpy warnings when broadcasting ``DataFrame`` to ``Series`` with comparison ops (:issue:`16378`, :issue:`16306`) | |||
| - Bug in ``Series`` with dtype='category' (:issue:`16524`) | |||
There was a problem hiding this comment.
Bug in Series construction with dtype='category'
pandas/tests/series/test_dtypes.py
Outdated
| @@ -248,3 +248,11 @@ def test_intercept_astype_object(self): | |||
|
|
|||
| result = df.values.squeeze() | |||
| assert (result[:, 0] == expected.values).all() | |||
|
|
|||
| def test_series_to_categorical(self): | |||
| series = Series(['a', 'b', 'c']) | |||
There was a problem hiding this comment.
add comment for the issue number
pandas/tests/series/test_dtypes.py
Outdated
| def test_series_to_categorical(self): | ||
| series = Series(['a', 'b', 'c']) | ||
|
|
||
| categorical_series = Series(series, dtype='category') |
pandas/tests/series/test_dtypes.py
Outdated
| categorical_series = Series(series, dtype='category') | ||
| expected = Series(['a', 'b', 'c'], dtype='category') | ||
|
|
||
| tm.assert_series_equal(categorical_series, expected) |
There was a problem hiding this comment.
categorical_series -> result
pandas/core/categorical.py
Outdated
| @@ -248,7 +248,8 @@ class Categorical(PandasObject): | |||
| __array_priority__ = 1000 | |||
| _typ = 'categorical' | |||
|
|
|||
| def __init__(self, values, categories=None, ordered=False, fastpath=False): | |||
| def __init__(self, values, categories=None, ordered=False, fastpath=False, | |||
There was a problem hiding this comment.
this fix should not be here.
instead in core/internals.py/Block/_astype.py
raise_one_error can be a named argument in the signature and not passed to the constructor
80ed6da to
b37fba3
Compare
|
@TomAugspurger @jreback I have made the changes according to your comments. |
|
@jreback hey I have made changes according to your previous comments, could you review the changes once. Thank You. |
|
@preddy5 I rebased and moved the release note to 0.20.3 (releasing in a couple weeks). I'll let Jeff comment on the changes, but I think we're good. |
doc/source/whatsnew/v0.20.3.txt
Outdated
| @@ -44,6 +44,7 @@ Conversion | |||
| ^^^^^^^^^^ | |||
|
|
|||
| - Bug in pickle compat prior to the v0.20.x series, when ``UTC`` is a timezone in a Series/DataFrame/Index (:issue:`16608`) | |||
| - Bug in ``Series`` with ``dtype='category'`` (:issue:`16524`) | |||
There was a problem hiding this comment.
be a bit more descriptive.
Bug in Series construction when passing type=category
|
@preddy5 Thanks! |
fix pandas-dev#16524 (cherry picked from commit 9fdea65)
Patch to bug fix #16524