Backport PR #57556 on branch 2.2.x (Remove PyArrow deprecation warning) (#57568)

Author: phofl

.github/workflows/unit-tests.yml

@@ -92,10 +92,7 @@ jobs:
           - name: "Numpy Dev"
             env_file: actions-311-numpydev.yaml
             pattern: "not slow and not network and not single_cpu"
-            # Currently restricted the warnings that error to Deprecation Warnings from numpy
-            # done since pyarrow isn't compatible with numpydev always
-            # TODO: work with pyarrow to revert this?
-            test_args: "-W error::DeprecationWarning:numpy -W error::FutureWarning:numpy"
+            test_args: "-W error::DeprecationWarning -W error::FutureWarning"
           - name: "Pyarrow Nightly"
             env_file: actions-311-pyarrownightly.yaml
             pattern: "not slow and not network and not single_cpu"

doc/source/whatsnew/v2.2.1.rst

@@ -67,6 +67,16 @@ Bug fixes
 
 Other
 ~~~~~
+
+.. note::
+
+    The ``DeprecationWarning`` that was raised when pandas was imported without PyArrow being
+    installed has been removed. This decision was made because the warning was too noisy for too
+    many users and a lot of feedback was collected about the decision to make PyArrow a required
+    dependency. Pandas is currently considering the decision whether or not PyArrow should be added
+    as a hard dependency in 3.0. Interested users can follow the discussion
+    `here <https://github.com/pandas-dev/pandas/issues/57073>`_.
+
 - Added the argument ``skipna`` to :meth:`DataFrameGroupBy.first`, :meth:`DataFrameGroupBy.last`, :meth:`SeriesGroupBy.first`, and :meth:`SeriesGroupBy.last`; achieving ``skipna=False`` used to be available via :meth:`DataFrameGroupBy.nth`, but the behavior was changed in pandas 2.0.0 (:issue:`57019`)
 - Added the argument ``skipna`` to :meth:`Resampler.first`, :meth:`Resampler.last` (:issue:`57019`)
 

pandas/__init__.py

@@ -203,36 +203,7 @@
         stacklevel=2,
     )
 
-# DeprecationWarning for missing pyarrow
-from pandas.compat.pyarrow import pa_version_under10p1, pa_not_found
-
-if pa_version_under10p1:
-    # pyarrow is either too old or nonexistent, warn
-    from pandas.compat._optional import VERSIONS
-
-    if pa_not_found:
-        pa_msg = "was not found to be installed on your system."
-    else:
-        pa_msg = (
-            f"was too old on your system - pyarrow {VERSIONS['pyarrow']} "
-            "is the current minimum supported version as of this release."
-        )
-
-    warnings.warn(
-        f"""
-Pyarrow will become a required dependency of pandas in the next major release of pandas (pandas 3.0),
-(to allow more performant data types, such as the Arrow string type, and better interoperability with other libraries)
-but {pa_msg}
-If this would cause problems for you,
-please provide us feedback at https://github.com/pandas-dev/pandas/issues/54466
-        """,  # noqa: E501
-        DeprecationWarning,
-        stacklevel=2,
-    )
-    del VERSIONS, pa_msg
-
-# Delete all unnecessary imported modules
-del pa_version_under10p1, pa_not_found, warnings, os
+del warnings, os
 
 # module level doc-string
 __doc__ = """

pandas/compat/pyarrow.py

@@ -8,7 +8,6 @@
     import pyarrow as pa
 
     _palv = Version(Version(pa.__version__).base_version)
-    pa_not_found = False
     pa_version_under10p1 = _palv < Version("10.0.1")
     pa_version_under11p0 = _palv < Version("11.0.0")
     pa_version_under12p0 = _palv < Version("12.0.0")
@@ -17,7 +16,6 @@
     pa_version_under14p1 = _palv < Version("14.0.1")
     pa_version_under15p0 = _palv < Version("15.0.0")
 except ImportError:
-    pa_not_found = True
     pa_version_under10p1 = True
     pa_version_under11p0 = True
     pa_version_under12p0 = True

pandas/tests/test_common.py

@@ -8,8 +8,6 @@
 import numpy as np
 import pytest
 
-import pandas.util._test_decorators as td
-
 import pandas as pd
 from pandas import Series
 import pandas._testing as tm
@@ -267,26 +265,3 @@ def test_bz2_missing_import():
     code = textwrap.dedent(code)
     call = [sys.executable, "-c", code]
     subprocess.check_output(call)
-
-
-@td.skip_if_installed("pyarrow")
-@pytest.mark.parametrize("module", ["pandas", "pandas.arrays"])
-def test_pyarrow_missing_warn(module):
-    # GH56896
-    response = subprocess.run(
-        [sys.executable, "-c", f"import {module}"],
-        capture_output=True,
-        check=True,
-    )
-    msg = """
-Pyarrow will become a required dependency of pandas in the next major release of pandas (pandas 3.0),
-(to allow more performant data types, such as the Arrow string type, and better interoperability with other libraries)
-but was not found to be installed on your system.
-If this would cause problems for you,
-please provide us feedback at https://github.com/pandas-dev/pandas/issues/54466
-"""  # noqa: E501
-    stderr_msg = response.stderr.decode("utf-8")
-    # Split by \n to avoid \r\n vs \n differences on Windows/Unix
-    # https://stackoverflow.com/questions/11989501/replacing-r-n-with-n
-    stderr_msg = "\n".join(stderr_msg.splitlines())
-    assert msg in stderr_msg