Function from_items

python/ray/data/read_api.py:166–254

Create a :class:`~ray.data.Dataset` from a list of local Python objects. Use this method to create small datasets from data that fits in memory. The column name defaults to "item". Examples: >>> import ray >>> ds = ray.data.from_items([1, 2, 3, 4, 5]) >>> ds #

(
    items: List[Any],
    *,
    parallelism: int = -1,
    override_num_blocks: Optional[int] = None,
)

@PublicAPI
def from_items(
    items: List[Any],
    *,
    parallelism: int = -1,
    override_num_blocks: Optional[int] = None,
) -> MaterializedDataset:
    """Create a :class:`~ray.data.Dataset` from a list of local Python objects.

    Use this method to create small datasets from data that fits in memory. The column
    name defaults to "item".

    Examples:

        >>> import ray
        >>> ds = ray.data.from_items([1, 2, 3, 4, 5])
        >>> ds  # doctest: +ELLIPSIS
        shape: (5, 1)
        ╭───────╮
        │ item  │
        │ ---   │
        │ int64 │
        ╞═══════╡
        │ 1     │
        │ 2     │
        │ 3     │
        │ 4     │
        │ 5     │
        ╰───────╯
        (Showing 5 of 5 rows)
        >>> ds.schema()
        Column  Type
        ------  ----
        item    int64

    Args:
        items: List of local Python objects.
        parallelism: This argument is deprecated. Use ``override_num_blocks`` argument.
        override_num_blocks: Override the number of output blocks from all read tasks.
            By default, the number of output blocks is dynamically decided based on
            input data size and available resources. You shouldn't manually set this
            value in most cases.

    Returns:
        A :class:`~ray.data.Dataset` holding the items.
    """
    import builtins

    parallelism = _get_num_output_blocks(parallelism, override_num_blocks)
    if parallelism == 0:
        raise ValueError(f"parallelism must be -1 or > 0, got: {parallelism}")

    detected_parallelism, _, _ = _autodetect_parallelism(
        parallelism,
        ray.util.get_current_placement_group(),
        DataContext.get_current(),
    )
    # Truncate parallelism to number of items to avoid empty blocks.
    detected_parallelism = min(len(items), detected_parallelism)