TST/REF: collect base tests by method (pandas-dev#37549)

Author: jbrockmendel

pandas/tests/base/test_drop_duplicates.py

@@ -5,15 +5,10 @@
 
 from pandas.compat import IS64, PYPY
 
-from pandas.core.dtypes.common import (
-    is_categorical_dtype,
-    is_datetime64_dtype,
-    is_datetime64tz_dtype,
-    is_object_dtype,
-)
+from pandas.core.dtypes.common import is_categorical_dtype, is_object_dtype
 
 import pandas as pd
-from pandas import DataFrame, Index, IntervalIndex, Series
+from pandas import DataFrame, Index, Series
 
 
 @pytest.mark.parametrize(
@@ -42,54 +37,6 @@ def test_binary_ops_docstring(klass, op_name, op):
     assert expected_str in getattr(klass, "r" + op_name).__doc__
 
 
-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")
-
-    # bug brought up by #1079
-    # changed from TypeError in 0.17.0
-    series[0] = np.nan
-
-    # noinspection PyComparisonWithNone
-    result = series == None  # noqa
-    assert not result.iat[0]
-    assert not result.iat[1]
-
-    # noinspection PyComparisonWithNone
-    result = series != None  # noqa
-    assert result.iat[0]
-    assert result.iat[1]
-
-    result = None == series  # noqa
-    assert not result.iat[0]
-    assert not result.iat[1]
-
-    result = None != series  # noqa
-    assert result.iat[0]
-    assert result.iat[1]
-
-    if is_datetime64_dtype(series.dtype) or is_datetime64tz_dtype(series.dtype):
-        # Following DatetimeIndex (and Timestamp) convention,
-        # inequality comparisons with Series[datetime64] raise
-        msg = "Invalid comparison"
-        with pytest.raises(TypeError, match=msg):
-            None > series
-        with pytest.raises(TypeError, match=msg):
-            series > None
-    else:
-        result = None > series
-        assert not result.iat[0]
-        assert not result.iat[1]
-
-        result = series < None
-        assert not result.iat[0]
-        assert not result.iat[1]
-
-
 def test_ndarray_compat_properties(index_or_series_obj):
     obj = index_or_series_obj
 

pandas/tests/base/test_misc.py

@@ -1,9 +1,10 @@
+from datetime import datetime
 import re
 
 import numpy as np
 import pytest
 
-from pandas import DataFrame
+from pandas import DataFrame, NaT
 import pandas._testing as tm
 
 
@@ -434,3 +435,27 @@ def test_drop_duplicates_null_in_object_column(nulls_fixture):
     df = DataFrame([[1, nulls_fixture], [2, "a"]], dtype=object)
     result = df.drop_duplicates()
     tm.assert_frame_equal(result, df)
+
+
+@pytest.mark.parametrize("keep", ["first", "last", False])
+def test_drop_duplicates_series_vs_dataframe(keep):
+    # GH#14192
+    df = DataFrame(
+        {
+            "a": [1, 1, 1, "one", "one"],
+            "b": [2, 2, np.nan, np.nan, np.nan],
+            "c": [3, 3, np.nan, np.nan, "three"],
+            "d": [1, 2, 3, 4, 4],
+            "e": [
+                datetime(2015, 1, 1),
+                datetime(2015, 1, 1),
+                datetime(2015, 2, 1),
+                NaT,
+                NaT,
+            ],
+        }
+    )
+    for column in df.columns:
+        dropped_frame = df[[column]].drop_duplicates(keep=keep)
+        dropped_series = df[column].drop_duplicates(keep=keep)
+        tm.assert_frame_equal(dropped_frame, dropped_series.to_frame())

pandas/tests/frame/methods/test_drop_duplicates.py

@@ -1,6 +1,7 @@
 import numpy as np
 
-from pandas import DataFrame
+from pandas import DataFrame, MultiIndex
+import pandas._testing as tm
 from pandas.core.arrays import PandasArray
 
 
@@ -50,3 +51,17 @@ def test_to_dict_of_blocks_item_cache():
     assert df.loc[0, "b"] == "foo"
 
     assert df["b"] is ser
+
+
+def test_set_change_dtype_slice():
+    # GH#8850
+    cols = MultiIndex.from_tuples([("1st", "a"), ("2nd", "b"), ("3rd", "c")])
+    df = DataFrame([[1.0, 2, 3], [4.0, 5, 6]], columns=cols)
+    df["2nd"] = df["2nd"] * 2.0
+
+    blocks = df._to_dict_of_blocks()
+    assert sorted(blocks.keys()) == ["float64", "int64"]
+    tm.assert_frame_equal(
+        blocks["float64"], DataFrame([[1.0, 4.0], [4.0, 10.0]], columns=cols[:2])
+    )
+    tm.assert_frame_equal(blocks["int64"], DataFrame([[3], [6]], columns=cols[2:]))

pandas/tests/frame/methods/test_to_dict_of_blocks.py

@@ -8,7 +8,7 @@
 from pandas._libs.internals import BlockPlacement
 
 import pandas as pd
-from pandas import Categorical, DataFrame, DatetimeIndex, Index, MultiIndex, Series
+from pandas import Categorical, DataFrame, DatetimeIndex, Index, Series
 import pandas._testing as tm
 import pandas.core.algorithms as algos
 from pandas.core.arrays import DatetimeArray, SparseArray, TimedeltaArray
@@ -1152,20 +1152,6 @@ def test_missing_unicode_key():
         df.loc[:, "\u05d0"]  # should not raise UnicodeEncodeError
 
 
-def test_set_change_dtype_slice():
-    # GH#8850
-    cols = MultiIndex.from_tuples([("1st", "a"), ("2nd", "b"), ("3rd", "c")])
-    df = DataFrame([[1.0, 2, 3], [4.0, 5, 6]], columns=cols)
-    df["2nd"] = df["2nd"] * 2.0
-
-    blocks = df._to_dict_of_blocks()
-    assert sorted(blocks.keys()) == ["float64", "int64"]
-    tm.assert_frame_equal(
-        blocks["float64"], DataFrame([[1.0, 4.0], [4.0, 10.0]], columns=cols[:2])
-    )
-    tm.assert_frame_equal(blocks["int64"], DataFrame([[3], [6]], columns=cols[2:]))
-
-
 def test_single_block_manager_fastpath_deprecated():
     # GH#33092
     ser = Series(range(3))

pandas/tests/internals/test_internals.py

@@ -7,8 +7,18 @@
 
 from pandas._libs.tslibs import IncompatibleFrequency
 
+from pandas.core.dtypes.common import is_datetime64_dtype, is_datetime64tz_dtype
+
 import pandas as pd
-from pandas import Categorical, Index, Series, bdate_range, date_range, isna
+from pandas import (
+    Categorical,
+    Index,
+    IntervalIndex,
+    Series,
+    bdate_range,
+    date_range,
+    isna,
+)
 import pandas._testing as tm
 from pandas.core import nanops, ops
 
@@ -779,3 +789,51 @@ def test_binop_maybe_preserve_name(self, datetime_series):
     def test_scalarop_preserve_name(self, datetime_series):
         result = datetime_series * 2
         assert result.name == datetime_series.name
+
+
+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")
+
+    # bug brought up by #1079
+    # changed from TypeError in 0.17.0
+    series[0] = np.nan
+
+    # noinspection PyComparisonWithNone
+    result = series == None  # noqa
+    assert not result.iat[0]
+    assert not result.iat[1]
+
+    # noinspection PyComparisonWithNone
+    result = series != None  # noqa
+    assert result.iat[0]
+    assert result.iat[1]
+
+    result = None == series  # noqa
+    assert not result.iat[0]
+    assert not result.iat[1]
+
+    result = None != series  # noqa
+    assert result.iat[0]
+    assert result.iat[1]
+
+    if is_datetime64_dtype(series.dtype) or is_datetime64tz_dtype(series.dtype):
+        # Following DatetimeIndex (and Timestamp) convention,
+        # inequality comparisons with Series[datetime64] raise
+        msg = "Invalid comparison"
+        with pytest.raises(TypeError, match=msg):
+            None > series
+        with pytest.raises(TypeError, match=msg):
+            series > None
+    else:
+        result = None > series
+        assert not result.iat[0]
+        assert not result.iat[1]
+
+        result = series < None
+        assert not result.iat[0]
+        assert not result.iat[1]