"""Tests whether the given value is an ndarray (and not a TF tensor/var)."""
# TODO(tomhennigan) Support __array_interface__ too.
return hasattr(value, "__array__") and not (
isinstance(value, ops.Tensor)
or isinstance(value, resource_variable_ops.BaseResourceVariable)
or hasattr(value, "_should_act_as_resource_variable")
# For legacy reasons we do not automatically promote Numpy strings.
or isinstance(value, np.str_)
# NumPy dtypes have __array__ as unbound methods.
or isinstance(value, type)
# CompositeTensors should be flattened instead.
or isinstance(value, composite_tensor.CompositeTensor))
_pywrap_utils.RegisterType("GenericType", GenericType)
class OrderedCollectionType(trace.TraceType):
"""Represents an ordered collection of TraceType objects.
Attributes:
components: The sequence of TraceType objects that this class represents.
"""
def __init__(self, *components: trace.TraceType):
self.components = components
def _has_same_structure(self, other):
if not isinstance(other, type(self)):
return False
`CompositeTensor`s; use `CompositeTensor._type_spec` instead.
Args:
type_object: A Python `type` object representing the type of values
accepted by `converter_fn`.
converter_fn: A function that takes one argument (an instance of the
type represented by `type_object`) and returns a `TypeSpec`.
allow_subclass: If true, then use `isinstance(value, type_object)` to
check for matches. If false, then use `type(value) is type_object`.
"""
_, type_object = tf_decorator.unwrap(type_object)
_TYPE_CONVERSION_FUNCTION_REGISTRY.append(
(type_object, converter_fn, allow_subclass))
_pywrap_utils.RegisterType("TypeSpec", TypeSpec)
_TYPE_SPEC_TO_NAME = {}
_NAME_TO_TYPE_SPEC = {}
# Regular expression for valid TypeSpec names.
_REGISTERED_NAME_RE = re.compile(r"^(\w+\.)+\w+$")
# TODO(b/173744905) tf_export this as "tf.register_type_spec". (And add a
# usage example to the docstring, once the API is public.)
#
# TODO(b/173744905) Update this decorator to apply to ExtensionType rather than
# TypeSpec (once we do refactoring to move to_components/from_components from
device=iterator_resource.device)
]
super(_IteratorSaveable, self).__init__(iterator_resource, specs, name)
def restore(self, restored_tensors, restored_shapes):
with ops.colocate_with(self.op):
return gen_dataset_ops.deserialize_iterator(
self.op, restored_tensors[0])
@deprecation.deprecated(
None, "Use `tf.data.Iterator.get_next_as_optional()` instead.")
@tf_export("data.experimental.get_next_as_optional")
def get_next_as_optional(iterator):
"""Returns a `tf.experimental.Optional` with the next element of the iterator.
If the iterator has reached the end of the sequence, the returned
`tf.experimental.Optional` will have no value.
Args:
iterator: A `tf.data.Iterator`.
Returns:
A `tf.experimental.Optional` object which either contains the next element
of the iterator (if it exists) or no value.
"""
return iterator.get_next_as_optional()
_pywrap_utils.RegisterType("OwnedIterator", OwnedIterator)
if not isinstance(other, WeakrefType):
return False
if self._object() is None or other._object() is None:
return False
if self._object() is other._object():
return True
return self._object == other._object
def __hash__(self):
return self._object_hash
_pywrap_utils.RegisterType("GenericType", GenericType)
_pywrap_utils.RegisterType("WeakrefType", WeakrefType)
class OrderedCollectionType(trace.TraceType):
"""Represents an ordered collection of TraceType objects.
Attributes:
components: The sequence of TraceType objects that this class represents.
"""
def __init__(self, *components: trace.TraceType):
self.components = components
def _has_same_structure(self, other):
if not isinstance(other, type(self)):
return False
s = "BoundedTensorSpec(shape={}, dtype={}, name={}, minimum={}, maximum={})"
return s.format(self.shape, repr(self.dtype), repr(self.name),
repr(self.minimum), repr(self.maximum))
def __eq__(self, other):
tensor_spec_eq = super(BoundedTensorSpec, self).__eq__(other)
return (tensor_spec_eq and np.allclose(self.minimum, other.minimum)
and np.allclose(self.maximum, other.maximum))
def __hash__(self):
return hash((self._shape, self.dtype))
def __reduce__(self):
return BoundedTensorSpec, (self._shape, self._dtype, self._minimum,
self._maximum, self._name)
def _serialize(self):
return (self._shape, self._dtype, self._minimum, self._maximum,
self._name)
trace_type.register_serializable(BoundedTensorSpec)
_pywrap_utils.RegisterType("TensorSpec", TensorSpec)
# Note: we do not include Tensor names when constructing TypeSpecs.
type_spec.register_type_spec_from_value_converter(
ops.Tensor, lambda tensor: TensorSpec(tensor.shape, tensor.dtype))
type_spec.register_type_spec_from_value_converter(
np.ndarray, lambda array: TensorSpec(array.shape, array.dtype))
# must be the shape of a vector.
if shape.ndims is not None and shape.ndims != 1:
raise ValueError(
f"Expected a shape with 1 dimension. Obtained: {shape} "
f"which has {shape.ndims} dimensions.")
rank = tensor_shape.dimension_value(shape[0])
return SparseTensorSpec(tensor_shape.unknown_shape(rank), self.dtype)
def consumers(self):
return self._consumers()
SparseTensorValue = collections.namedtuple(
"SparseTensorValue", ["indices", "values", "dense_shape"])
tf_export(v1=["SparseTensorValue"])(SparseTensorValue)
_pywrap_utils.RegisterType("SparseTensorValue", SparseTensorValue)
@tf_export("SparseTensorSpec")
@type_spec.register("tf.SparseTensorSpec")
class SparseTensorSpec(type_spec.BatchableTypeSpec):
"""Type specification for a `tf.sparse.SparseTensor`."""
__slots__ = ["_shape", "_dtype"]
value_type = property(lambda self: SparseTensor)
def __init__(self, shape=None, dtype=dtypes.float32):
"""Constructs a type specification for a `tf.sparse.SparseTensor`.
Args:
def _convert_variables_to_tensors(self):
"""Converts ResourceVariable components to Tensors.
Override this method to explicitly convert ResourceVariables embedded in the
CompositeTensor to Tensors. By default, it returns the CompositeTensor
unchanged.
Returns:
A CompositeTensor with all its ResourceVariable components converted to
Tensors.
"""
return self
_pywrap_utils.RegisterType("CompositeTensor", CompositeTensor)
def replace_composites_with_components(structure):
"""Recursively replaces CompositeTensors with their components.
Args:
structure: A `nest`-compatible structure, possibly containing composite
tensors.
Returns:
A copy of `structure`, where each composite tensor has been replaced by
its components. The result will contain no composite tensors.
Note that `nest.flatten(replace_composites_with_components(structure))`
returns the same value as `nest.flatten(structure)`.
"""
