ak.values_astype
----------------

.. py:module: ak.values_astype

Defined in `awkward.operations.ak_values_astype <https://github.com/scikit-hep/awkward/blob/21af5dccad55a8b9e447b7be4af8b35226f1f95c/src/awkward/operations/ak_values_astype.py>`__ on `line 16 <https://github.com/scikit-hep/awkward/blob/21af5dccad55a8b9e447b7be4af8b35226f1f95c/src/awkward/operations/ak_values_astype.py#L16>`__.

.. py:function:: ak.values_astype(array, to, *, including_unknown=False, highlevel=True, behavior=None, attrs=None)


    :param array: Array-like data (anything :py:obj:`ak.to_layout` recognizes).
    :param to: Type to convert the numbers into.
    :type to: dtype or dtype specifier
    :param including_unknown: If True, the ``unknown`` type is considered
                          a value type and is converted to the specified dtype; if False,
                          ``unknown`` will remain ``unknown``.
    :type including_unknown: bool
    :param highlevel: If True, return an :py:obj:`ak.Array`; otherwise, return
                  a low-level :py:obj:`ak.contents.Content` subclass.
    :type highlevel: bool
    :param behavior: Custom :py:obj:`ak.behavior` for the output array, if
                 high-level.
    :type behavior: None or dict
    :param attrs: Custom attributes for the output array, if
              high-level.
    :type attrs: None or dict

Converts all numbers in the array to a new type, leaving the structure
untouched.

For example,

.. code-block:: python


    >>> array = ak.Array([1.1, 2.2, 3.3, 4.4, 5.5])
    >>> ak.values_astype(array, np.int32)
    <Array [1, 2, 3, 4, 5] type='5 * int32'>

and

.. code-block:: python


    >>> array = ak.Array([[1.1, 2.2, 3.3], [], [4.4, 5.5]])
    >>> ak.values_astype(array, np.int32)
    <Array [[1, 2, 3], [], [4, 5]] type='3 * var * int32'>

Note, when converting values to a ``np.datetime64`` type that is unitless, a
default '[us]' unit is assumed - until further specified as numpy dtypes.

For example,

.. code-block:: python


    >>> array = ak.Array([1567416600000])
    >>> ak.values_astype(array, "datetime64[ms]")
    <Array [2019-09-02T09:30:00.000] type='1 * datetime64[ms]'>

or

.. code-block:: python


    >>> array = ak.Array([1567416600000])
    >>> ak.values_astype(array, np.dtype("M8[ms]"))
    <Array [2019-09-02T09:30:00.000] type='1 * datetime64[ms]['>

See also :py:obj:`ak.strings_astype`.