cythonarrays.array_descriptors module

class cythonarrays.array_descriptors.ArrayDescriptor(name, dtype, ndim, shape=None, dtype_numpy=None, default=None)

Bases: object

describes an array used as an instance attribute of a cython class

Parameters:

  • name (str)

  • dtype (str)

  • ndim (int)

  • shape (Tuple[int | str])

  • dtype_numpy (str)

__init__(name, dtype, ndim, shape=None, dtype_numpy=None, default=None)
Parameters:

  • name (str) – the name of the array

  • dtype (str) – the numpy-dtype of the array for internal storage

  • ndim (int) – the number of dimensions of the array

  • shape (Tuple[Union[int, str]]) – the shape of the array

  • dtype_numpy (str) – the numpy-dtype of the array that should be returned

  • default (number) – the default value

get_shape(instance=None)

get the shape of the array defined by the ArrayDescriptor using the information from the ArrayShape-instance

Parameters:

instance (ArrayShapes) – the CDefClass-instance holding the information on the allowed shapes

Return type:

Tuple[int]

Returns:

a tuple with the shape

static to_dim(arr, ndim)

Casts the array to ndim dimensions

Parameters:

ndim (int) – if self.ndim > ndim and the shape of the first axis <> 0: raise ValueError

Return type:

ndarray

Returns:

array of ndim dimensions

validate_array(arr, instance=None)

checks if the ndim (and the shape, if specified) match and casts the array to the dtype specified

Parameters:

  • arr (ndarray) – the array to test

  • instance (ArrayShapes) – the CDefClass-instance holding the information on the allowed shapes

Return type:

ndarray

Returns:

the validated array casted to the target dimension and dtype

property dtype_numpy: str

return the numpy dtype

property shape: Tuple[int | str]

the shape of the array described, with numbers or strings

Parameters:

value (Union[int, str, bytes, Tuple[Union[int, str]]]) – the setter can receive the shape provided as a number, a string (comma separated for several dimensions) or a tuple/list of strings or ints