TST: Added match argument for most uses of tm.assert_produces_warning (#58396)

* Fix for all FutureWarnings * Add match for most warnings * Cleaner code --------- Co-authored-by: Abdulaziz Aloqeely <52792999+DAzVise@users.noreply.github.com>

Author: Aloqeely

doc/source/development/contributing_codebase.rst

@@ -557,11 +557,12 @@ is being raised, using ``pytest.raises`` instead.
 Testing a warning
 ^^^^^^^^^^^^^^^^^
 
-Use ``tm.assert_produces_warning`` as a context manager to check that a block of code raises a warning.
+Use ``tm.assert_produces_warning`` as a context manager to check that a block of code raises a warning
+and specify the warning message using the ``match`` argument.
 
 .. code-block:: python
 
- with tm.assert_produces_warning(DeprecationWarning):
+ with tm.assert_produces_warning(DeprecationWarning, match="the warning message"):
  pd.deprecated_function()
 
 If a warning should specifically not happen in a block of code, pass ``False`` into the context manager.

pandas/tests/arrays/sparse/test_constructors.py

@@ -90,13 +90,13 @@ def test_constructor_warns_when_losing_timezone(self):
  dti = pd.date_range("2016-01-01", periods=3, tz="US/Pacific")
 
  expected = SparseArray(np.asarray(dti, dtype="datetime64[ns]"))
-
- with tm.assert_produces_warning(UserWarning):
+ msg = "loses timezone information"
+ with tm.assert_produces_warning(UserWarning, match=msg):
  result = SparseArray(dti)
 
  tm.assert_sp_array_equal(result, expected)
 
- with tm.assert_produces_warning(UserWarning):
+ with tm.assert_produces_warning(UserWarning, match=msg):
  result = SparseArray(pd.Series(dti))
 
  tm.assert_sp_array_equal(result, expected)

pandas/tests/arrays/test_datetimelike.py

@@ -778,7 +778,7 @@ def test_to_period_2d(self, arr1d):
  arr2d = arr1d.reshape(1, -1)
 
  warn = None if arr1d.tz is None else UserWarning
- with tm.assert_produces_warning(warn):
+ with tm.assert_produces_warning(warn, match="will drop timezone information"):
  result = arr2d.to_period("D")
  expected = arr1d.to_period("D").reshape(1, -1)
  tm.assert_period_array_equal(result, expected)

pandas/tests/computation/test_eval.py

@@ -1014,7 +1014,8 @@ def test_performance_warning_for_poor_alignment(
  else:
  seen = False
 
- with tm.assert_produces_warning(seen):
+ msg = "Alignment difference on axis 1 is larger than an order of magnitude"
+ with tm.assert_produces_warning(seen, match=msg):
  pd.eval("df + s", engine=engine, parser=parser)
 
  s = Series(np.random.default_rng(2).standard_normal(1000))
@@ -1036,7 +1037,7 @@ def test_performance_warning_for_poor_alignment(
  else:
  wrn = False
 
- with tm.assert_produces_warning(wrn) as w:
+ with tm.assert_produces_warning(wrn, match=msg) as w:
  pd.eval("df + s", engine=engine, parser=parser)
 
  if not is_python_engine and performance_warning:

pandas/tests/dtypes/test_common.py

@@ -797,5 +797,5 @@ def test_pandas_dtype_numpy_warning():
 
 def test_pandas_dtype_ea_not_instance():
  # GH 31356 GH 54592
- with tm.assert_produces_warning(UserWarning):
+ with tm.assert_produces_warning(UserWarning, match="without any arguments"):
  assert pandas_dtype(CategoricalDtype) == CategoricalDtype()

pandas/tests/dtypes/test_generic.py

@@ -124,7 +124,7 @@ def test_setattr_warnings():
  # this should not raise a warning
  df.two.not_an_index = [1, 2]
 
- with tm.assert_produces_warning(UserWarning):
+ with tm.assert_produces_warning(UserWarning, match="doesn't allow columns"):
  # warn when setting column to nonexistent name
  df.four = df.two + 2
  assert df.four.sum() > df.two.sum()

pandas/tests/frame/indexing/test_indexing.py

@@ -145,7 +145,7 @@ def test_getitem_boolean(self, mixed_float_frame, mixed_int_frame, datetime_fram
  # we are producing a warning that since the passed boolean
  # key is not the same as the given index, we will reindex
  # not sure this is really necessary
- with tm.assert_produces_warning(UserWarning):
+ with tm.assert_produces_warning(UserWarning, match="will be reindexed"):
  indexer_obj = indexer_obj.reindex(datetime_frame.index[::-1])
  subframe_obj = datetime_frame[indexer_obj]
  tm.assert_frame_equal(subframe_obj, subframe)

pandas/tests/frame/indexing/test_setitem.py

@@ -711,7 +711,10 @@ def test_setitem_npmatrix_2d(self):
  df["np-array"] = a
 
  # Instantiation of `np.matrix` gives PendingDeprecationWarning
- with tm.assert_produces_warning(PendingDeprecationWarning):
+ with tm.assert_produces_warning(
+ PendingDeprecationWarning,
+ match="matrix subclass is not the recommended way to represent matrices",
+ ):
  df["np-matrix"] = np.matrix(a)
 
  tm.assert_frame_equal(df, expected)

pandas/tests/frame/methods/test_to_dict.py

@@ -166,7 +166,7 @@ def test_to_dict_not_unique_warning(self):
  # GH#16927: When converting to a dict, if a column has a non-unique name
  # it will be dropped, throwing a warning.
  df = DataFrame([[1, 2, 3]], columns=["a", "a", "b"])
- with tm.assert_produces_warning(UserWarning):
+ with tm.assert_produces_warning(UserWarning, match="columns will be omitted"):
  df.to_dict()
 
  @pytest.mark.filterwarnings("ignore::UserWarning")

pandas/tests/frame/test_arithmetic.py

@@ -1097,25 +1097,24 @@ def test_binop_other(self, op, value, dtype, switch_numexpr_min_elements):
  and expr.USE_NUMEXPR
  and switch_numexpr_min_elements == 0
  ):
- warn = UserWarning # "evaluating in Python space because ..."
+ warn = UserWarning
  else:
  msg = (
  f"cannot perform __{op.__name__}__ with this "
  "index type: (DatetimeArray|TimedeltaArray)"
  )
 
  with pytest.raises(TypeError, match=msg):
- with tm.assert_produces_warning(warn):
+ with tm.assert_produces_warning(warn, match="evaluating in Python"):
  op(df, elem.value)
 
  elif (op, dtype) in skip:
  if op in [operator.add, operator.mul]:
  if expr.USE_NUMEXPR and switch_numexpr_min_elements == 0:
- # "evaluating in Python space because ..."
  warn = UserWarning
  else:
  warn = None
- with tm.assert_produces_warning(warn):
+ with tm.assert_produces_warning(warn, match="evaluating in Python"):
  op(df, elem.value)
 
  else: