def assert_series_equal(cls, left, right, *args, **kwargs):
# base class tests hard-code expected values with numpy dtypes,
# whereas we generally want the corresponding PandasDtype
if (isinstance(right, pd.Series)
and not isinstance(right.dtype, ExtensionDtype)
and isinstance(left.dtype, PandasDtype)):
right = right.astype(PandasDtype(right.dtype))
return tm.assert_series_equal(left, right, *args, **kwargs)
def test_fillna_fill_other(self, data_missing):
# Same as the parent class test, but with PandasDtype for expected["B"]
# instead of equivalent numpy dtype
data = data_missing
result = pd.DataFrame({"A": data, "B": [np.nan] * len(data)}).fillna({"B": 0.0})
expected = pd.DataFrame({"A": data, "B": [0.0] * len(result)})
expected["B"] = expected["B"].astype(PandasDtype(expected["B"].dtype))
self.assert_frame_equal(result, expected)
def __init__(self, values: np.ndarray | PandasArray, copy: bool = False):
if isinstance(values, type(self)):
values = values._ndarray
if not isinstance(values, np.ndarray):
raise ValueError(
f"'values' must be a NumPy array, not {type(values).__name__}")
if values.ndim == 0:
# Technically we support 2, but do not advertise that fact.
raise ValueError("PandasArray must be 1-dimensional.")
if copy:
values = values.copy()
dtype = PandasDtype(values.dtype)
super().__init__(values, dtype)
def dtype(request):
return PandasDtype(np.dtype(request.param))
def test_dtype_univalent(any_numpy_dtype):
dtype = PandasDtype(any_numpy_dtype)
result = PandasDtype(dtype)
assert result == dtype
def test_constructor_from_string():
result = PandasDtype.construct_from_string("int64")
expected = PandasDtype(np.dtype("int64"))
assert result == expected
def test_repr():
dtype = PandasDtype(np.dtype("int64"))
assert repr(dtype) == "PandasDtype('int64')"
def test_is_boolean(dtype, expected):
dtype = PandasDtype(dtype)
assert dtype._is_boolean is expected
def test_is_numeric(dtype, expected):
dtype = PandasDtype(dtype)
assert dtype._is_numeric is expected
