BUG: Increased support for subclassed types.

doc/source/whatsnew/v1.1.0.rst

@@ -748,10 +748,13 @@ Reshaping
 - Bug in :meth:`concat` where when passing a non-dict mapping as ``objs`` would raise a ``TypeError`` (:issue:`32863`)
 - :meth:`DataFrame.agg` now provides more descriptive ``SpecificationError`` message when attempting to aggregating non-existant column (:issue:`32755`)
 - Bug in :meth:`DataFrame.unstack` when MultiIndexed columns and MultiIndexed rows were used (:issue:`32624`, :issue:`24729` and :issue:`28306`)
+- Bug in :meth:`DataFrame.corrwith()`, :meth:`DataFrame.memory_usage()`, :meth:`DataFrame.dot()`,
+  :meth:`DataFrame.idxmin()`, :meth:`DataFrame.idxmax()`, :meth:`DataFrame.duplicated()`, :meth:`DataFrame.isin()`,
+  :meth:`DataFrame.count()`, :meth:`Series.explode()`, :meth:`Series.asof()` and :meth:`DataFrame.asof()` not
+  returning subclassed types. (:issue:`31331`)
 - Bug in :func:`concat` was not allowing for concatenation of ``DataFrame`` and ``Series`` with duplicate keys (:issue:`33654`)
 - Bug in :func:`cut` raised an error when non-unique labels (:issue:`33141`)
 
-
 Sparse
 ^^^^^^
 - Creating a :class:`SparseArray` from timezone-aware dtype will issue a warning before dropping timezone information, instead of doing so silently (:issue:`32501`)

pandas/conftest.py

@@ -709,6 +709,17 @@ def all_boolean_reductions(request):
     return request.param
 
 
+_all_reductions = _all_numeric_reductions + _all_boolean_reductions
+
+
+@pytest.fixture(params=_all_reductions)
+def all_reductions(request):
+    """
+    Fixture for all (boolean + numeric) reduction names.
+    """
+    return request.param
+
+
 @pytest.fixture(params=["__eq__", "__ne__", "__le__", "__lt__", "__ge__", "__gt__"])
 def all_compare_operators(request):
     """

pandas/core/frame.py

@@ -1173,13 +1173,13 @@ def dot(self, other):
                 np.dot(lvals, rvals), index=left.index, columns=other.columns
             )
         elif isinstance(other, Series):
-            return Series(np.dot(lvals, rvals), index=left.index)
+            return self._constructor_sliced(np.dot(lvals, rvals), index=left.index)
         elif isinstance(rvals, (np.ndarray, Index)):
             result = np.dot(lvals, rvals)
             if result.ndim == 2:
                 return self._constructor(result, index=left.index)
             else:
-                return Series(result, index=left.index)
+                return self._constructor_sliced(result, index=left.index)
         else:  # pragma: no cover
             raise TypeError(f"unsupported type: {type(other)}")
 
@@ -2533,14 +2533,14 @@ def memory_usage(self, index=True, deep=False) -> Series:
         >>> df['object'].astype('category').memory_usage(deep=True)
         5216
         """
-        result = Series(
+        result = self._constructor_sliced(
             [c.memory_usage(index=False, deep=deep) for col, c in self.items()],
             index=self.columns,
         )
         if index:
-            result = Series(self.index.memory_usage(deep=deep), index=["Index"]).append(
-                result
-            )
+            result = self._constructor_sliced(
+                self.index.memory_usage(deep=deep), index=["Index"]
+            ).append(result)
         return result
 
     def transpose(self, *args, copy: bool = False) -> "DataFrame":
@@ -5013,7 +5013,7 @@ def duplicated(
         from pandas._libs.hashtable import duplicated_int64, _SIZE_HINT_LIMIT
 
         if self.empty:
-            return Series(dtype=bool)
+            return self._constructor_sliced(dtype=bool)
 
         def f(vals):
             labels, shape = algorithms.factorize(
@@ -5045,7 +5045,7 @@ def f(vals):
         labels, shape = map(list, zip(*map(f, vals)))
 
         ids = get_group_index(labels, shape, sort=False, xnull=False)
-        return Series(duplicated_int64(ids, keep), index=self.index)
+        return self._constructor_sliced(duplicated_int64(ids, keep), index=self.index)
 
     # ----------------------------------------------------------------------
     # Sorting
@@ -8084,7 +8084,7 @@ def corrwith(self, other, axis=0, drop=False, method="pearson") -> Series:
             def c(x):
                 return nanops.nancorr(x[0], x[1], method=method)
 
-            correl = Series(
+            correl = self._constructor_sliced(
                 map(c, zip(left.values.T, right.values.T)), index=left.columns
             )
 
@@ -8197,7 +8197,7 @@ def count(self, axis=0, level=None, numeric_only=False):
 
         # GH #423
         if len(frame._get_axis(axis)) == 0:
-            result = Series(0, index=frame._get_agg_axis(axis))
+            result = self._constructor_sliced(0, index=frame._get_agg_axis(axis))
         else:
             if frame._is_mixed_type or frame._mgr.any_extension_types:
                 # the or any_extension_types is really only hit for single-
@@ -8207,7 +8207,9 @@ def count(self, axis=0, level=None, numeric_only=False):
                 # GH13407
                 series_counts = notna(frame).sum(axis=axis)
                 counts = series_counts.values
-                result = Series(counts, index=frame._get_agg_axis(axis))
+                result = self._constructor_sliced(
+                    counts, index=frame._get_agg_axis(axis)
+                )
 
         return result.astype("int64")
 
@@ -8250,9 +8252,9 @@ def _count_level(self, level, axis=0, numeric_only=False):
         counts = lib.count_level_2d(mask, level_codes, len(level_index), axis=axis)
 
         if axis == 1:
-            result = DataFrame(counts, index=agg_axis, columns=level_index)
+            result = self._constructor(counts, index=agg_axis, columns=level_index)
         else:
-            result = DataFrame(counts, index=level_index, columns=agg_axis)
+            result = self._constructor(counts, index=level_index, columns=agg_axis)
 
         return result
 
@@ -8523,7 +8525,7 @@ def idxmin(self, axis=0, skipna=True) -> Series:
 
         index = self._get_axis(axis)
         result = [index[i] if i >= 0 else np.nan for i in indices]
-        return Series(result, index=self._get_agg_axis(axis))
+        return self._constructor_sliced(result, index=self._get_agg_axis(axis))
 
     def idxmax(self, axis=0, skipna=True) -> Series:
         """
@@ -8596,7 +8598,7 @@ def idxmax(self, axis=0, skipna=True) -> Series:
 
         index = self._get_axis(axis)
         result = [index[i] if i >= 0 else np.nan for i in indices]
-        return Series(result, index=self._get_agg_axis(axis))
+        return self._constructor_sliced(result, index=self._get_agg_axis(axis))
 
     def _get_agg_axis(self, axis_num: int) -> Index:
         """
@@ -8940,7 +8942,7 @@ def isin(self, values) -> "DataFrame":
                     "to be passed to DataFrame.isin(), "
                     f"you passed a '{type(values).__name__}'"
                 )
-            return DataFrame(
+            return self._constructor(
                 algorithms.isin(self.values.ravel(), values).reshape(self.shape),
                 self.index,
                 self.columns,

pandas/core/generic.py

@@ -7067,9 +7067,9 @@ def asof(self, where, subset=None):
 
             if where < start:
                 if not is_series:
-                    from pandas import Series
-
-                    return Series(index=self.columns, name=where, dtype=np.float64)
+                    return self._constructor_sliced(
+                        index=self.columns, name=where, dtype=np.float64
+                    )
                 return np.nan
 
             # It's always much faster to use a *while* loop here for
@@ -7096,13 +7096,11 @@ def asof(self, where, subset=None):
             if is_series:
                 return self._constructor(np.nan, index=where, name=self.name)
             elif is_list:
-                from pandas import DataFrame
-
-                return DataFrame(np.nan, index=where, columns=self.columns)
+                return self._constructor(np.nan, index=where, columns=self.columns)
             else:
-                from pandas import Series
-
-                return Series(np.nan, index=self.columns, name=where[0])
+                return self._constructor_sliced(
+                    np.nan, index=self.columns, name=where[0]
+                )
 
         locs = self.index.asof_locs(where, ~(nulls._values))
 

pandas/core/series.py

@@ -3625,7 +3625,9 @@ def explode(self) -> "Series":
 
         values, counts = reshape.explode(np.asarray(self.array))
 
-        result = Series(values, index=self.index.repeat(counts), name=self.name)
+        result = self._constructor(
+            values, index=self.index.repeat(counts), name=self.name
+        )
         return result
 
     def unstack(self, level=-1, fill_value=None):

pandas/tests/frame/test_subclass.py

@@ -1,6 +1,8 @@
 import numpy as np
 import pytest
 
+import pandas.util._test_decorators as td
+
 import pandas as pd
 from pandas import DataFrame, Index, MultiIndex, Series
 import pandas._testing as tm
@@ -560,16 +562,123 @@ def strech(row):
         assert not isinstance(result, tm.SubclassedDataFrame)
         tm.assert_series_equal(result, expected)
 
-    def test_subclassed_numeric_reductions(self, all_numeric_reductions):
+    def test_subclassed_reductions(self, all_reductions):
         # GH 25596
 
         df = tm.SubclassedDataFrame({"A": [1, 2, 3], "B": [4, 5, 6], "C": [7, 8, 9]})
-        result = getattr(df, all_numeric_reductions)()
+        result = getattr(df, all_reductions)()
         assert isinstance(result, tm.SubclassedSeries)
 
-    def test_subclassed_boolean_reductions(self, all_boolean_reductions):
-        # GH 25596
+    def test_subclassed_count(self):
+
+        df = tm.SubclassedDataFrame(
+            {
+                "Person": ["John", "Myla", "Lewis", "John", "Myla"],
+                "Age": [24.0, np.nan, 21.0, 33, 26],
+                "Single": [False, True, True, True, False],
+            }
+        )
+        result = df.count()
+        assert isinstance(result, tm.SubclassedSeries)
+
+        df = tm.SubclassedDataFrame({"A": [1, 0, 3], "B": [0, 5, 6], "C": [7, 8, 0]})
+        result = df.count()
+        assert isinstance(result, tm.SubclassedSeries)
+
+        df = tm.SubclassedDataFrame(
+            [[10, 11, 12, 13], [20, 21, 22, 23], [30, 31, 32, 33], [40, 41, 42, 43]],
+            index=MultiIndex.from_tuples(
+                list(zip(list("AABB"), list("cdcd"))), names=["aaa", "ccc"]
+            ),
+            columns=MultiIndex.from_tuples(
+                list(zip(list("WWXX"), list("yzyz"))), names=["www", "yyy"]
+            ),
+        )
+        result = df.count(level=1)
+        assert isinstance(result, tm.SubclassedDataFrame)
+
+        df = tm.SubclassedDataFrame()
+        result = df.count()
+        assert isinstance(result, tm.SubclassedSeries)
+
+    def test_isin(self):
+
+        df = tm.SubclassedDataFrame(
+            {"num_legs": [2, 4], "num_wings": [2, 0]}, index=["falcon", "dog"]
+        )
+        result = df.isin([0, 2])
+        assert isinstance(result, tm.SubclassedDataFrame)
+
+    def test_duplicated(self):
 
         df = tm.SubclassedDataFrame({"A": [1, 2, 3], "B": [4, 5, 6], "C": [7, 8, 9]})
-        result = getattr(df, all_boolean_reductions)()
+        result = df.duplicated()
+        assert isinstance(result, tm.SubclassedSeries)
+
+        df = tm.SubclassedDataFrame()
+        result = df.duplicated()
+        assert isinstance(result, tm.SubclassedSeries)
+
+    @pytest.mark.parametrize("idx_method", ["idxmax", "idxmin"])
+    def test_idx(self, idx_method):
+
+        df = tm.SubclassedDataFrame({"A": [1, 2, 3], "B": [4, 5, 6], "C": [7, 8, 9]})
+        result = getattr(df, idx_method)()
+        assert isinstance(result, tm.SubclassedSeries)
+
+    def test_dot(self):
+
+        df = tm.SubclassedDataFrame([[0, 1, -2, -1], [1, 1, 1, 1]])
+        s = tm.SubclassedSeries([1, 1, 2, 1])
+        result = df.dot(s)
+        assert isinstance(result, tm.SubclassedSeries)
+
+        df = tm.SubclassedDataFrame([[0, 1, -2, -1], [1, 1, 1, 1]])
+        s = tm.SubclassedDataFrame([1, 1, 2, 1])
+        result = df.dot(s)
+        assert isinstance(result, tm.SubclassedDataFrame)
+
+    def test_memory_usage(self):
+
+        df = tm.SubclassedDataFrame({"A": [1, 2, 3], "B": [4, 5, 6], "C": [7, 8, 9]})
+        result = df.memory_usage()
+        assert isinstance(result, tm.SubclassedSeries)
+
+        result = df.memory_usage(index=False)
+        assert isinstance(result, tm.SubclassedSeries)
+
+    @td.skip_if_no_scipy
+    def test_corrwith(self):
+        index = ["a", "b", "c", "d", "e"]
+        columns = ["one", "two", "three", "four"]
+        df1 = tm.SubclassedDataFrame(
+            np.random.randn(5, 4), index=index, columns=columns
+        )
+        df2 = tm.SubclassedDataFrame(
+            np.random.randn(4, 4), index=index[:4], columns=columns
+        )
+        correls = df1.corrwith(df2, axis=1, drop=True, method="kendall")
+
+        assert isinstance(correls, (tm.SubclassedSeries))
+
+    def test_asof(self):
+
+        N = 3
+        rng = pd.date_range("1/1/1990", periods=N, freq="53s")
+        df = tm.SubclassedDataFrame(
+            {
+                "A": [np.nan, np.nan, np.nan],
+                "B": [np.nan, np.nan, np.nan],
+                "C": [np.nan, np.nan, np.nan],
+            },
+            index=rng,
+        )
+
+        result = df.asof(rng[-2:])
+        assert isinstance(result, tm.SubclassedDataFrame)
+
+        result = df.asof(rng[-2])
+        assert isinstance(result, tm.SubclassedSeries)
+
+        result = df.asof("1989-12-31")
         assert isinstance(result, tm.SubclassedSeries)

pandas/tests/series/test_subclass.py

@@ -1,3 +1,6 @@
+import numpy as np
+
+import pandas as pd
 import pandas._testing as tm
 
 
@@ -35,3 +38,16 @@ def test_subclass_empty_repr(self):
         with tm.assert_produces_warning(DeprecationWarning, check_stacklevel=False):
             sub_series = tm.SubclassedSeries()
         assert "SubclassedSeries" in repr(sub_series)
+
+    def test_asof(self):
+        N = 3
+        rng = pd.date_range("1/1/1990", periods=N, freq="53s")
+        s = tm.SubclassedSeries({"A": [np.nan, np.nan, np.nan]}, index=rng)
+
+        result = s.asof(rng[-2:])
+        assert isinstance(result, tm.SubclassedSeries)
+
+    def test_explode(self):
+        s = tm.SubclassedSeries([[1, 2, 3], "foo", [], [3, 4]])
+        result = s.explode()
+        assert isinstance(result, tm.SubclassedSeries)