def _pandas(cls, column, type_list, **kwargs):
comp_types = []
for type_ in type_list:
try:
comp_types.append(np.dtype(type_).type)
except TypeError:
try:
pd_type = getattr(pd, type_)
if isinstance(pd_type, type):
comp_types.append(pd_type)
except AttributeError:
pass
try:
pd_type = getattr(pd.core.dtypes.dtypes, type_)
if isinstance(pd_type, type):
comp_types.append(pd_type)
except AttributeError:
pass
native_type = _native_type_type_map(type_)
if native_type is not None:
comp_types.extend(native_type)
if len(comp_types) < 1:
raise ValueError("No recognized numpy/python type in list: %s" %
type_list)
return column.map(lambda x: isinstance(x, tuple(comp_types)))
def _validate_pandas(
self, actual_column_type, expected_types_list,
):
if expected_types_list is None:
success = True
else:
comp_types = []
for type_ in expected_types_list:
try:
comp_types.append(np.dtype(type_).type)
except TypeError:
try:
pd_type = getattr(pd, type_)
if isinstance(pd_type, type):
comp_types.append(pd_type)
except AttributeError:
pass
try:
pd_type = getattr(pd.core.dtypes.dtypes, type_)
if isinstance(pd_type, type):
comp_types.append(pd_type)
except AttributeError:
pass
native_type = _native_type_type_map(type_)
if native_type is not None:
comp_types.extend(native_type)
success = actual_column_type in comp_types
return {
"success": success,
"result": {"observed_value": actual_column_type.type.__name__},
}
def _validate_pandas(
self,
actual_column_type,
expected_types_list,
):
if expected_types_list is None:
success = True
else:
comp_types = []
for type_ in expected_types_list:
try:
comp_types.append(np.dtype(type_).type)
comp_types.append(np.dtype(type_))
except TypeError:
try:
pd_type = getattr(pd, type_)
except AttributeError:
pass
else:
if isinstance(pd_type, type):
comp_types.append(pd_type)
try:
if isinstance(
pd_type(),
pd.core.dtypes.base.ExtensionDtype):
comp_types.append(pd_type())
except TypeError:
pass
try:
pd_type = getattr(pd.core.dtypes.dtypes, type_)
if isinstance(pd_type, type):
comp_types.append(pd_type)
except AttributeError:
pass
native_type = _native_type_type_map(type_)
if native_type is not None:
comp_types.extend(native_type)
# TODO: Remove when Numpy >=1.21 is pinned as a dependency
_pandas_supports_extension_dtypes = version.parse(
pd.__version__) >= version.parse("0.24")
_numpy_doesnt_support_extensions_properly = version.parse(
np.__version__) < version.parse("1.21")
if (_numpy_doesnt_support_extensions_properly
and _pandas_supports_extension_dtypes):
# This works around a bug where Pandas nullable int types aren't compatible with Numpy dtypes
# Note: Can't do set difference, the whole bugfix is because numpy types can't be compared to
# ExtensionDtypes
actual_type_is_ext_dtype = isinstance(
actual_column_type, pd.core.dtypes.base.ExtensionDtype)
comp_types = {
dtype
for dtype in comp_types
if isinstance(dtype, pd.core.dtypes.base.ExtensionDtype) ==
actual_type_is_ext_dtype
}
###
success = actual_column_type in comp_types
return {
"success": success,
"result": {
"observed_value": actual_column_type.type.__name__
},
}
