TST: Reduce deep copies in the test suite (#62970)

Author: mroeschke

pandas/conftest.py

@@ -722,7 +722,7 @@ def index(request):
         - ...
     """
     # copy to avoid mutation, e.g. setting .name
-    return indices_dict[request.param].copy()
+    return indices_dict[request.param].copy(deep=False)
 
 
 @pytest.fixture(
@@ -735,7 +735,7 @@ def index_flat(request):
     index fixture, but excluding MultiIndex cases.
     """
     key = request.param
-    return indices_dict[key].copy()
+    return indices_dict[key].copy(deep=False)
 
 
 @pytest.fixture(
@@ -758,18 +758,15 @@ def index_with_missing(request):
 
     MultiIndex is excluded because isna() is not defined for MultiIndex.
     """
-
-    # GH 35538. Use deep copy to avoid illusive bug on np-dev
-    # GHA pipeline that writes into indices_dict despite copy
-    ind = indices_dict[request.param].copy(deep=True)
-    vals = ind.values.copy()
+    ind = indices_dict[request.param]
     if request.param in ["tuples", "mi-with-dt64tz-level", "multi"]:
         # For setting missing values in the top level of MultiIndex
         vals = ind.tolist()
         vals[0] = (None,) + vals[0][1:]
         vals[-1] = (None,) + vals[-1][1:]
         return MultiIndex.from_tuples(vals)
     else:
+        vals = ind.values.copy()
         vals[0] = None
         vals[-1] = None
         return type(ind)(vals)
@@ -850,7 +847,7 @@ def index_or_series_obj(request):
     Fixture for tests on indexes, series and series with a narrow dtype
     copy to avoid mutation, e.g. setting .name
     """
-    return _index_or_series_objs[request.param].copy(deep=True)
+    return _index_or_series_objs[request.param].copy(deep=False)
 
 
 _typ_objects_series = {
@@ -873,7 +870,7 @@ def index_or_series_memory_obj(request):
     series with empty objects type
     copy to avoid mutation, e.g. setting .name
     """
-    return _index_or_series_memory_objs[request.param].copy(deep=True)
+    return _index_or_series_memory_objs[request.param].copy(deep=False)
 
 
 # ----------------------------------------------------------------

pandas/tests/arrays/interval/test_interval.py

@@ -115,9 +115,7 @@ def test_shift_datetime(self):
 class TestSetitem:
     def test_set_na(self, left_right_dtypes):
         left, right = left_right_dtypes
-        left = left.copy(deep=True)
-        right = right.copy(deep=True)
-        result = IntervalArray.from_arrays(left, right)
+        result = IntervalArray.from_arrays(left, right, copy=True)
 
         if result.dtype.subtype.kind not in ["m", "M"]:
             msg = "'value' should be an interval type, got <.*NaTType'> instead."
@@ -168,8 +166,6 @@ def test_setitem_mismatched_closed(self):
 class TestReductions:
     def test_min_max_invalid_axis(self, left_right_dtypes):
         left, right = left_right_dtypes
-        left = left.copy(deep=True)
-        right = right.copy(deep=True)
         arr = IntervalArray.from_arrays(left, right)
 
         msg = "`axis` must be fewer than the number of dimensions"
@@ -188,8 +184,6 @@ def test_min_max_invalid_axis(self, left_right_dtypes):
     def test_min_max(self, left_right_dtypes, index_or_series_or_array):
         # GH#44746
         left, right = left_right_dtypes
-        left = left.copy(deep=True)
-        right = right.copy(deep=True)
         arr = IntervalArray.from_arrays(left, right)
 
         # The expected results below are only valid if monotonic

pandas/tests/base/test_fillna.py

@@ -44,6 +44,7 @@ def test_fillna_null(null_obj, index_or_series_obj):
     elif isinstance(obj, MultiIndex):
         pytest.skip(f"MultiIndex can't hold '{null_obj}'")
 
+    obj = obj.copy(deep=True)
     values = obj._values
     fill_value = values[0]
     expected = values.copy()

pandas/tests/base/test_unique.py

@@ -46,6 +46,7 @@ def test_unique_null(null_obj, index_or_series_obj, using_nan_is_na):
     ):
         pytest.skip("NaN is not a valid NA for this dtype.")
 
+    obj = obj.copy(deep=True)
     values = obj._values
     values[0:2] = null_obj
 
@@ -87,6 +88,7 @@ def test_nunique_null(null_obj, index_or_series_obj):
     elif isinstance(obj, pd.MultiIndex):
         pytest.skip(f"MultiIndex can't hold '{null_obj}'")
 
+    obj = obj.copy(deep=True)
     values = obj._values
     values[0:2] = null_obj
 

pandas/tests/base/test_value_counts.py

@@ -55,15 +55,15 @@ def test_value_counts(index_or_series_obj):
 @pytest.mark.filterwarnings(r"ignore:PeriodDtype\[B\] is deprecated:FutureWarning")
 def test_value_counts_null(null_obj, index_or_series_obj):
     orig = index_or_series_obj
-    obj = orig.copy()
 
-    if not allow_na_ops(obj):
+    if not allow_na_ops(orig):
         pytest.skip("type doesn't allow for NA operations")
-    elif len(obj) < 1:
+    elif len(orig) < 1:
         pytest.skip("Test doesn't make sense on empty data")
     elif isinstance(orig, MultiIndex):
         pytest.skip(f"MultiIndex can't hold '{null_obj}'")
 
+    obj = orig.copy(deep=True)
     values = obj._values
     values[0:2] = null_obj
 

pandas/tests/copy_view/test_methods.py

@@ -560,7 +560,7 @@ def test_to_frame():
     ],
     ids=["shallow-copy", "reset_index", "rename", "select_dtypes"],
 )
-def test_chained_methods(request, method, idx):
+def test_chained_methods(method, idx):
     df = DataFrame({"a": [1, 2, 3], "b": [4, 5, 6], "c": [0.1, 0.2, 0.3]})
     df_orig = df.copy()
 

pandas/tests/indexes/test_common.py

@@ -231,7 +231,7 @@ def test_unique(self, index_flat):
         if not index._can_hold_na:
             pytest.skip("Skip na-check if index cannot hold na")
 
-        vals = index._values[[0] * 5]
+        vals = index._values.copy()[[0] * 5]
         vals[0] = np.nan
 
         vals_unique = vals[:2]