Function from_array

dask/array/_array_expr/_collection.py:1005–1195

Create dask array from something that looks like an array. Input must have a ``.shape``, ``.ndim``, ``.dtype`` and support numpy-style slicing. Parameters ---------- x : array_like chunks : int, tuple How to chunk the array. Must be one of the following forms:

(
    x,
    chunks="auto",
    lock=False,
    asarray=None,
    fancy=True,
    getitem=None,
    meta=None,
    inline_array=False,
    name=None,
)

def from_array(
    x,
    chunks="auto",
    lock=False,
    asarray=None,
    fancy=True,
    getitem=None,
    meta=None,
    inline_array=False,
    name=None,
):
    """Create dask array from something that looks like an array.

    Input must have a ``.shape``, ``.ndim``, ``.dtype`` and support numpy-style slicing.

    Parameters
    ----------
    x : array_like
    chunks : int, tuple
        How to chunk the array. Must be one of the following forms:

        - A blocksize like 1000.
        - A blockshape like (1000, 1000).
        - Explicit sizes of all blocks along all dimensions like
          ((1000, 1000, 500), (400, 400)).
        - A size in bytes, like "100 MiB" which will choose a uniform
          block-like shape
        - The word "auto" which acts like the above, but uses a configuration
          value ``array.chunk-size`` for the chunk size

        -1 or None as a blocksize indicate the size of the corresponding
        dimension.
    name : str or bool, optional
        The key name to use for the array. Defaults to a hash of ``x``.

        Hashing is useful if the same value of ``x`` is used to create multiple
        arrays, as Dask can then recognise that they're the same and
        avoid duplicate computations. However, it can also be slow, and if the
        array is not contiguous it is copied for hashing. If the array uses
        stride tricks (such as :func:`numpy.broadcast_to` or
        :func:`skimage.util.view_as_windows`) to have a larger logical
        than physical size, this copy can cause excessive memory usage.

        If you don't need the deduplication provided by hashing, use
        ``name=False`` to generate a random name instead of hashing, which
        avoids the pitfalls described above. Using ``name=True`` is
        equivalent to the default.

        By default, hashing uses python's standard sha1. This behaviour can be
        changed by installing cityhash, xxhash or murmurhash. If installed,
        a large-factor speedup can be obtained in the tokenisation step.

        .. note::

           Because this ``name`` is used as the key in task graphs, you should
           ensure that it uniquely identifies the data contained within. If
           you'd like to provide a descriptive name that is still unique, combine
           the descriptive name with :func:`dask.base.tokenize` of the