Module numpyctypes

Returns a ctypes structure of the array 'a' containing the arrays info (data, shape, strides, ndim). A check is made to ensure that the array has the specified dtype and requirements.

Example:

>>> myArray = np.arange(10.0)
>>> myCstruct = c_ndarray(myArray, dtype=np.double, ndim = 3, shape = (4,3,2), 
...                       requirements = ['c_contiguous'])

Parameters:

• a (ndarray) - the numpy array to be converted
• dtype (numpy dtype) - the required dtype of the array, convert if it doesn't match
• ndim (integer) - the required number of axes of the array, complain if it doesn't match
• shape (tuple) - required shape of the array, complain if it doesn't match
• requirements (list) - "ensurearray", "aligned", "fortran", "f_contiguous", or "c_contiguous". Convert if it doesn't match.

Returns: ctypes structure

ctypes structure with the fields:

• data: pointer to the data : the type is determined with the dtype of the array, and with ctypesDict.
• shape: pointer to long array : size of each of the dimensions
• strides: pointer to long array : strides in elements (not bytes)

ctypesDict

Value:

{'d': C.c_double, 'b': C.c_char, 'h': C.c_short, 'i': C.c_int, 'l': C. c_long, 'q': C.c_longlong, 'B': C.c_ubyte, 'H': C.c_ushort, 'I': C.c_u int, 'L': C.c_ulong, 'Q': C.c_ulonglong}