CLN: TODOs (#44733)

Author: jbrockmendel

pandas/tests/arrays/boolean/test_construction.py

@@ -321,19 +321,3 @@ def test_to_numpy_copy():
     result = arr.to_numpy(dtype=bool, copy=True)
     result[0] = False
     tm.assert_extension_array_equal(arr, pd.array([True, False, True], dtype="boolean"))
-
-
-# FIXME: don't leave commented out
-# TODO when BooleanArray coerces to object dtype numpy array, need to do conversion
-# manually in the indexing code
-# def test_indexing_boolean_mask():
-#     arr = pd.array([1, 2, 3, 4], dtype="Int64")
-#     mask = pd.array([True, False, True, False], dtype="boolean")
-#     result = arr[mask]
-#     expected = pd.array([1, 3], dtype="Int64")
-#     tm.assert_extension_array_equal(result, expected)
-
-#     # missing values -> error
-#     mask = pd.array([True, False, True, None], dtype="boolean")
-#     with pytest.raises(IndexError):
-#         result = arr[mask]

pandas/tests/arrays/masked/test_arithmetic.py

@@ -7,7 +7,6 @@
 
 import pandas as pd
 import pandas._testing as tm
-from pandas.core.arrays import ExtensionArray
 
 # integer dtypes
 arrays = [pd.array([1, 2, 3, None], dtype=dtype) for dtype in tm.ALL_INT_EA_DTYPES]
@@ -71,11 +70,7 @@ def test_numpy_array_equivalence(data, all_arithmetic_operators):
 
     result = op(data, numpy_array)
     expected = op(data, pd_array)
-    if isinstance(expected, ExtensionArray):
-        tm.assert_extension_array_equal(result, expected)
-    else:
-        # TODO div still gives float ndarray -> remove this once we have Float EA
-        tm.assert_numpy_array_equal(result, expected)
+    tm.assert_extension_array_equal(result, expected)
 
 
 # Test equivalence with Series and DataFrame ops

pandas/tests/extension/test_sparse.py

@@ -328,11 +328,14 @@ def test_where_series(self, data, na_value):
         expected = pd.Series(cls._from_sequence([a, b, b, b], dtype=data.dtype))
         self.assert_series_equal(result, expected)
 
-    def test_combine_first(self, data):
+    def test_combine_first(self, data, request):
         if data.dtype.subtype == "int":
             # Right now this is upcasted to float, just like combine_first
             # for Series[int]
-            pytest.skip("TODO(SparseArray.__setitem__ will preserve dtype.")
+            mark = pytest.mark.xfail(
+                reason="TODO(SparseArray.__setitem__) will preserve dtype."
+            )
+            request.node.add_marker(mark)
         super().test_combine_first(data)
 
     def test_searchsorted(self, data_for_sorting, as_series):

pandas/tests/frame/methods/test_align.py

@@ -55,9 +55,7 @@ def test_align_float(self, float_frame):
         # test fill value
         join_idx = float_frame.index.join(other.index)
         diff_a = float_frame.index.difference(join_idx)
-        diff_b = other.index.difference(join_idx)
         diff_a_vals = af.reindex(diff_a).values
-        diff_b_vals = bf.reindex(diff_b).values
         assert (diff_a_vals == -1).all()
 
         af, bf = float_frame.align(other, join="right", axis=0)
@@ -74,12 +72,8 @@ def test_align_float(self, float_frame):
         # test fill value
         join_idx = float_frame.index.join(other.index)
         diff_a = float_frame.index.difference(join_idx)
-        diff_b = other.index.difference(join_idx)
         diff_a_vals = af.reindex(diff_a).values
 
-        # TODO(wesm): unused?
-        diff_b_vals = bf.reindex(diff_b).values  # noqa
-
         assert (diff_a_vals == -1).all()
 
         af, bf = float_frame.align(other, join="inner", axis=1)

pandas/tests/frame/methods/test_quantile.py

@@ -571,7 +571,7 @@ def test_quantile_item_cache(self, using_array_manager):
 
 class TestQuantileExtensionDtype:
     # TODO: tests for axis=1?
-    # TODO: empty case?  might as well do dt64 and td64 here too
+    # TODO: empty case?
 
     @pytest.fixture(
         params=[
@@ -581,6 +581,7 @@ class TestQuantileExtensionDtype:
             ),
             pd.period_range("2016-01-01", periods=9, freq="D"),
             pd.date_range("2016-01-01", periods=9, tz="US/Pacific"),
+            pd.timedelta_range("1 Day", periods=9),
             pd.array(np.arange(9), dtype="Int64"),
             pd.array(np.arange(9), dtype="Float64"),
         ],
@@ -650,7 +651,18 @@ def test_quantile_ea_with_na(self, obj, index):
 
     # TODO(GH#39763): filtering can be removed after GH#39763 is fixed
     @pytest.mark.filterwarnings("ignore:Using .astype to convert:FutureWarning")
-    def test_quantile_ea_all_na(self, obj, index, frame_or_series):
+    def test_quantile_ea_all_na(
+        self, obj, index, frame_or_series, using_array_manager, request
+    ):
+        if (
+            using_array_manager
+            and frame_or_series is DataFrame
+            and index.dtype == "m8[ns]"
+        ):
+            mark = pytest.mark.xfail(
+                reason="obj.astype fails bc obj is incorrectly dt64 at this point"
+            )
+            request.node.add_marker(mark)
 
         obj.iloc[:] = index._na_value
 

pandas/tests/groupby/test_apply.py

@@ -360,7 +360,7 @@ def test_apply_frame_not_as_index_column_name(df):
     grouped = df.groupby(["A", "B"], as_index=False)
     result = grouped.apply(len)
     expected = grouped.count().rename(columns={"C": np.nan}).drop(columns="D")
-    # TODO: Use assert_frame_equal when column name is not np.nan (GH 36306)
+    # TODO(GH#34306): Use assert_frame_equal when column name is not np.nan
     tm.assert_index_equal(result.index, expected.index)
     tm.assert_numpy_array_equal(result.values, expected.values)
 
@@ -1134,9 +1134,8 @@ def test_positional_slice_groups_datetimelike():
     tm.assert_frame_equal(result, expected)
 
 
-def test_doctest_example2():
+def test_groupby_apply_shape_cache_safety():
     # GH#42702 this fails if we cache_readonly Block.shape
-    # TODO: more informative name
     df = DataFrame({"A": ["a", "a", "b"], "B": [1, 2, 3], "C": [4, 6, 5]})
     gb = df.groupby("A")
     result = gb[["B", "C"]].apply(lambda x: x.astype(float).max() - x.min())

pandas/tests/indexes/datetimes/test_constructors.py

@@ -69,20 +69,20 @@ def test_shallow_copy_inherits_array_freq(self, index):
 
     def test_categorical_preserves_tz(self):
         # GH#18664 retain tz when going DTI-->Categorical-->DTI
-        # TODO: parametrize over DatetimeIndex/DatetimeArray
-        #  once pd.CategoricalIndex(DTA) works
-
         dti = DatetimeIndex(
             [pd.NaT, "2015-01-01", "1999-04-06 15:14:13", "2015-01-01"], tz="US/Eastern"
         )
 
-        ci = pd.CategoricalIndex(dti)
-        carr = pd.Categorical(dti)
-        cser = pd.Series(ci)
+        for dtobj in [dti, dti._data]:
+            # works for DatetimeIndex or DatetimeArray
+
+            ci = pd.CategoricalIndex(dtobj)
+            carr = pd.Categorical(dtobj)
+            cser = pd.Series(ci)
 
-        for obj in [ci, carr, cser]:
-            result = DatetimeIndex(obj)
-            tm.assert_index_equal(result, dti)
+            for obj in [ci, carr, cser]:
+                result = DatetimeIndex(obj)
+                tm.assert_index_equal(result, dti)
 
     def test_dti_with_period_data_raises(self):
         # GH#23675

pandas/tests/indexes/test_base.py

@@ -1364,9 +1364,9 @@ def test_str_to_bytes_raises(self):
     @pytest.mark.filterwarnings("ignore:elementwise comparison failed:FutureWarning")
     def test_index_with_tuple_bool(self):
         # GH34123
-        # TODO: remove tupleize_cols=False once correct behaviour is restored
         # TODO: also this op right now produces FutureWarning from numpy
-        idx = Index([("a", "b"), ("b", "c"), ("c", "a")], tupleize_cols=False)
+        #  https://github.com/numpy/numpy/issues/11521
+        idx = Index([("a", "b"), ("b", "c"), ("c", "a")])
         result = idx == ("c", "a")
         expected = np.array([False, False, True])
         tm.assert_numpy_array_equal(result, expected)

pandas/tests/indexes/test_setops.py

@@ -453,10 +453,9 @@ def test_intersection_difference_match_empty(self, index, sort):
 @pytest.mark.parametrize(
     "method", ["intersection", "union", "difference", "symmetric_difference"]
 )
-def test_setop_with_categorical(index, sort, method):
-    if isinstance(index, MultiIndex):  # TODO: flat_index?
-        # tested separately in tests.indexes.multi.test_setops
-        return
+def test_setop_with_categorical(index_flat, sort, method):
+    # MultiIndex tested separately in tests.indexes.multi.test_setops
+    index = index_flat
 
     other = index.astype("category")
     exact = "equiv" if isinstance(index, RangeIndex) else True

pandas/tests/indexing/test_chaining_and_caching.py

@@ -502,8 +502,8 @@ def test_iloc_setitem_chained_assignment(self):
 
             df["bb"].iloc[0] = 0.13
 
-            # TODO: unused
-            df_tmp = df.iloc[ck]  # noqa
+            # GH#3970 this lookup used to break the chained setting to 0.15
+            df.iloc[ck]
 
             df["bb"].iloc[0] = 0.15
             assert df["bb"].iloc[0] == 0.15

pandas/tests/indexing/test_indexing.py

@@ -232,11 +232,11 @@ def test_dups_fancy_indexing_across_dtypes(self):
         df.head()
         str(df)
         result = DataFrame([[1, 2, 1.0, 2.0, 3.0, "foo", "bar"]])
-        result.columns = list("aaaaaaa")
+        result.columns = list("aaaaaaa")  # GH#3468
 
-        # TODO(wesm): unused?
-        df_v = df.iloc[:, 4]  # noqa
-        res_v = result.iloc[:, 4]  # noqa
+        # GH#3509 smoke tests for indexing with duplicate columns
+        df.iloc[:, 4]
+        result.iloc[:, 4]
 
         tm.assert_frame_equal(df, result)
 

pandas/tests/indexing/test_loc.py

@@ -199,8 +199,8 @@ def test_loc_getitem_single_boolean_arg(self, obj, key, exp):
             assert res == exp
 
 
-class TestLoc2:
-    # TODO: better name, just separating out things that rely on base class
+class TestLocBaseIndependent:
+    # Tests for loc that do not depend on subclassing Base
     @pytest.mark.parametrize(
         "msg, key",
         [

pandas/tests/series/indexing/test_getitem.py

@@ -110,8 +110,8 @@ def test_getitem_int64(self, datetime_series):
         idx = np.int64(5)
         assert datetime_series[idx] == datetime_series[5]
 
-    # TODO: better name/GH ref?
-    def test_getitem_regression(self):
+    def test_getitem_full_range(self):
+        # github.com/pandas-dev/pandas/commit/4f433773141d2eb384325714a2776bcc5b2e20f7
         ser = Series(range(5), index=list(range(5)))
         result = ser[list(range(5))]
         tm.assert_series_equal(result, ser)
@@ -240,7 +240,6 @@ def test_getitem_partial_str_slice_high_reso_with_timedeltaindex(self):
         result = ser["1 days, 10:11:12.001001"]
         assert result == ser.iloc[1001]
 
-    # TODO: redundant with test_getitem_ndim_deprecated?
     def test_getitem_slice_2d(self, datetime_series):
         # GH#30588 multi-dimensional indexing deprecated
 

pandas/tests/series/test_arithmetic.py

@@ -16,7 +16,6 @@
 from pandas import (
     Categorical,
     Index,
-    IntervalIndex,
     Series,
     Timedelta,
     bdate_range,
@@ -828,9 +827,7 @@ def test_series_inplace_ops(self, dtype1, dtype2, dtype_expected, dtype_mul):
 
 def test_none_comparison(series_with_simple_index):
     series = series_with_simple_index
-    if isinstance(series.index, IntervalIndex):
-        # IntervalIndex breaks on "series[0] = np.nan" below
-        pytest.skip("IntervalIndex doesn't support assignment")
+
     if len(series) < 1:
         pytest.skip("Test doesn't make sense on empty data")