Backport PR pandas-dev#57551: BLD: Add pyarrow extra for pip installa…

…tion (pandas-dev#57557)

.github/workflows/package-checks.yml

@@ -24,7 +24,7 @@ jobs:
     runs-on: ubuntu-22.04
     strategy:
       matrix:
-        extra: ["test", "performance", "computation", "fss", "aws", "gcp", "excel", "parquet", "feather", "hdf5", "spss", "postgresql", "mysql", "sql-other", "html", "xml", "plot", "output-formatting", "clipboard", "compression", "consortium-standard", "all"]
+        extra: ["test", "pyarrow", "performance", "computation", "fss", "aws", "gcp", "excel", "parquet", "feather", "hdf5", "spss", "postgresql", "mysql", "sql-other", "html", "xml", "plot", "output-formatting", "clipboard", "compression", "consortium-standard", "all"]
       fail-fast: false
     name: Install Extras - ${{ matrix.extra }}
     concurrency:

doc/source/whatsnew/v2.2.1.rst

@@ -9,6 +9,12 @@ including other versions of pandas.
 {{ header }}
 
 .. ---------------------------------------------------------------------------
+.. _whatsnew_221.enhancements:
+
+Enhancements
+~~~~~~~~~~~~
+- Added ``pyarrow`` pip extra so users can install pandas and pyarrow with pip with ``pip install pandas[pyarrow]`` (:issue:`54466`)
+
 .. _whatsnew_221.regressions:
 
 Fixed regressions

pyproject.toml

@@ -64,6 +64,7 @@ matplotlib = "pandas:plotting._matplotlib"
 
 [project.optional-dependencies]
 test = ['hypothesis>=6.46.1', 'pytest>=7.3.2', 'pytest-xdist>=2.2.0']
+pyarrow = ['pyarrow>=10.0.1']
 performance = ['bottleneck>=1.3.6', 'numba>=0.56.4', 'numexpr>=2.8.4']
 computation = ['scipy>=1.10.0', 'xarray>=2022.12.0']
 fss = ['fsspec>=2022.11.0']