Save the date! Google I/O returns May 18-20 Register now

tf.contrib.framework.nest.assert_shallow_structure

View source on GitHub

Asserts that shallow_tree is a shallow structure of input_tree.

That is, this function tests if the input_tree structure can be created from the shallow_tree structure by replacing its leaf nodes with deeper tree structures.

Examples:

The following code will raise an exception:

  shallow_tree = {"a": "A", "b": "B"}
  input_tree = {"a": 1, "c": 2}
  assert_shallow_structure(shallow_tree, input_tree)

The following code will raise an exception:

  shallow_tree = ["a", "b"]
  input_tree = ["c", ["d", "e"], "f"]
  assert_shallow_structure(shallow_tree, input_tree)

The following code will not raise an exception:

  shallow_tree = ["a", "b"]
  input_tree = ["c", ["d", "e"], "f"]
  assert_shallow_structure(shallow_tree, input_tree,
    check_subtrees_length=False)

shallow_tree an arbitrarily nested structure.
input_tree an arbitrarily nested structure.
check_types if True (default) the sequence types of shallow_tree and input_tree have to be the same. Note that even with check_types==True, this function will consider two different namedtuple classes with the same name and _fields attribute to be the same class.
expand_composites If true, then composite tensors such as tf.SparseTensor and tf.RaggedTensor are expanded into their component tensors.
check_subtrees_length if True (default) the subtrees shallow_tree and input_tree have to be the same length. If False sequences are treated as key-value like mappings allowing them to be considered as valid subtrees. Note that this may drop parts of the input_tree.

TypeError If shallow_tree is a sequence but input_tree is not.
TypeError If the sequence types of shallow_tree are different from input_tree. Only raised if check_types is True.
ValueError If the sequence lengths of shallow_tree are different from input_tree.