Method unbatch

tensorflow/python/data/ops/dataset_ops.py:1592–1634

Splits elements of a dataset into multiple elements. For example, if elements of the dataset are shaped `[B, a0, a1, ...]`, where `B` may vary for each input element, then for each element in the dataset, the unbatched dataset will contain `B` consecutive elements of shape `[a0, a1,

(self)

            output_types=structure.get_flat_tensor_types(state_structure)))

  def unbatch(self):
    """Splits elements of a dataset into multiple elements.

    For example, if elements of the dataset are shaped `[B, a0, a1, ...]`,
    where `B` may vary for each input element, then for each element in the
    dataset, the unbatched dataset will contain `B` consecutive elements
    of shape `[a0, a1, ...]`.

    ```python
    # NOTE: The following example uses `{ ... }` to represent the contents
    # of a dataset.
    ds = { ['a', 'b', 'c'], ['a', 'b'], ['a', 'b', 'c', 'd'] }

    ds.unbatch() == {'a', 'b', 'c', 'a', 'b', 'a', 'b', 'c', 'd'}
    ```

    Returns:
      A `Dataset` transformation function, which can be passed to
      `tf.data.Dataset.apply`.
    """

    # NOTE(mrry): We must ensure that any non-tensor components in `dataset`
    # are normalized to their dense tensor representation, so that the
    # non-tensor oblivious unbatching logic will slice them appropriately.
    # This leads to a somewhat inefficient re-encoding step for all non-tensor
    # components.
    #
    # TODO(mrry): Consider optimizing this if it turns out to be a bottleneck.
    def normalize(arg, *rest):
      # pylint: disable=protected-access
      if rest:
        return structure.to_batched_tensor_list(self.element_spec,
                                                (arg,) + rest)
      else:
        return structure.to_batched_tensor_list(self.element_spec, arg)

    normalized_dataset = self.map(normalize)

    # NOTE(mrry): Our `map()` has lost information about the structure of
    # non-tensor components, so re-apply the structure of the original dataset.
    restructured_dataset = _RestructuredDataset(normalized_dataset,
                                                self.element_spec)
    return _UnbatchDataset(restructured_dataset)

  def with_options(self, options):
    """Returns a new `tf.data.Dataset` with the given options set.