RLS: 1.1.2

[simonjayhawkins]

Tracking issue for the 1.1.2 release.

https://github.com/pandas-dev/pandas/milestone/76

Please do not remove/change milestones from these issues without a note explaining the reasoning (changing milestones doesn't trigger notification)

During issue triage, regressions from 1.0.5 or 1.1.0 to 1.1.0 or 1.1.1 should be milestoned 1.1.2 in the first instance.

[simonjayhawkins]

new test failures on 32 bit builds (this test is also failing on nightly on the same builds that were already failing test_datetime64_factorize xref #35921

=========================== short test summary info ===========================
FAILED test_venv/lib/site-packages/pandas/tests/base/test_factorize.py::test_series_factorize_na_sentinel_none
= 1 failed, 64031 passed, 7376 skipped, 1049 xfailed, 6 xpassed, 14 warnings in 881.67s (0:14:41) =

https://dev.azure.com/pandas-dev/pandas-wheels/_build?definitionId=3&_a=summary

test_series_factorize_na_sentinel_none added in #35852

[simonjayhawkins]

release notes are not currently in sync,

• xref BUG: item_cache invalidation in get_numeric_data #35882 (comment)
• xref CLN: backport mpl warning fix to 1.1.2 #36155 (comment)

https://github.com/simonjayhawkins/pandas-release/runs/1080383769?check_suite_focus=true

diff --git a/doc/source/whatsnew/v1.1.2.rst b/doc/source/whatsnew/v1.1.2.rst
index a0e57eb17..1e946d325 100644
--- a/doc/source/whatsnew/v1.1.2.rst
+++ b/doc/source/whatsnew/v1.1.2.rst
@@ -40,6 +40,7 @@ Bug fixes
 - Bug in :class:`Series` constructor incorrectly raising a ``TypeError`` when passed an ordered set (:issue:`36044`)
 - Bug in :meth:`Series.dt.isocalendar` and :meth:`DatetimeIndex.isocalendar` that returned incorrect year for certain dates (:issue:`36032`)
 - Bug in :class:`DataFrame` indexing returning an incorrect :class:`Series` in some cases when the series has been altered and a cache not invalidated (:issue:`33675`)
+- Bug in :meth:`DataFrame.corr` causing subsequent indexing lookups to be incorrect (:issue:`35882`)
 
 .. ---------------------------------------------------------------------------
 
@@ -48,7 +49,6 @@ Bug fixes
 Other
 ~~~~~
 - :meth:`factorize` now supports ``na_sentinel=None`` to include NaN in the uniques of the values and remove ``dropna`` keyword which was unintentionally exposed to public facing API in 1.1 version from :meth:`factorize` (:issue:`35667`)
-- :meth:`DataFrame.plot` and meth:`Series.plot` raise ``UserWarning`` about usage of FixedFormatter and FixedLocator (:issue:`35684` and :issue:`35945`)```

[simonjayhawkins]

open PRs milestoned 1.1.2

https://github.com/pandas-dev/pandas/pulls?q=is%3Aopen+is%3Apr+milestone%3A1.1.2

please comment with issue number here if these should be included (otherwise will release tomorrow and mark PRs as changed requested with move of release note)

[simonjayhawkins]

• broken directive in https://pandas.pydata.org/pandas-docs/dev/whatsnew/v1.1.2.html#other

[simonjayhawkins]

tagging locally after merging and backporting #36182 and the following final pre-tag checks

• MacPython.pandas-wheels building test 1.1.x - DO NOT MERGE MacPython/pandas-wheels#97
• release readiness passing https://github.com/simonjayhawkins/pandas-release/actions/runs/244560644
• release notes checked https://pandas.pydata.org/pandas-docs/dev/whatsnew/v1.1.2.html
• ci checks https://github.com/pandas-dev/pandas/commits/1.1.x

[simonjayhawkins]

grr... test_groupby_rolling_index_changed added in #36152 not allowing 32bit wheels to build cc @jreback fix or skip

[jreback]

grr... test_groupby_rolling_index_changed added in #36152 not allowing wheels to build cc @jreback fix or skip

skip!

[jreback]

I guess have to revert the backport?

[simonjayhawkins]

can skip in tests on 32bit or skip in MacPython.

[jreback]

whatever works

[simonjayhawkins]

@mroeschke fixed this in #35875 which wasn't backported. so I think test_groupby_rolling_index_changed won't fail nightly. and adding @pytest.mark.xfail(not compat.IS64, reason="GH-35294") can be done directly against 1.1.x (that's how the other tests are currently marked on that branch)

also the option of seeing if #35875 will backport

[simonjayhawkins]

wheels now building after #36220 MacPython/pandas-wheels#97 (but have merged to 1.1.x since so will need to repeat check)

[simonjayhawkins]

#36147 is not green. proceeding with second pass of final checks (if #36147 gets sorted in meantime can abort checks)

will be tagging locally after

• MacPython.pandas-wheels building test 1.1.x - DO NOT MERGE MacPython/pandas-wheels#97
• release readiness passing https://github.com/simonjayhawkins/pandas-release/actions/runs/244783932
• release notes checked https://pandas.pydata.org/pandas-docs/dev/whatsnew/v1.1.2.html
• ci checks https://github.com/pandas-dev/pandas/commits/1.1.x

[simonjayhawkins]

starting the release now.

using process in https://github.com/simonjayhawkins/pandas-release/blob/release-1.1.1/windows.md for this release (so can't reuse tested sdist built in https://github.com/simonjayhawkins/pandas-release/actions/runs/244783932 this time round.)

[simonjayhawkins]

I'm going to upload docs from locally built this time, but am also building at https://github.com/simonjayhawkins/pandas-release/actions/runs/245044502

I think need to tidy the doc build script to upload the built docs and just do this on actions next time (between push tag and create release)

[simonjayhawkins]

docs uploaded and wheels on PyPI

conda build progressing slowly conda-forge/pandas-feedstock#91

[simonjayhawkins]

available on conda-forge