| |
- inline(unique_name, args=(), py_types=(), np_types=(), code=None, code_path=None, support_code=None, support_code_path=None, extension_kwargs={}, return_type=None)
- Inline C code in your python code.
Parameters:
unique_name : string
A unique string identifying this bit of code. This should be valid
to use as a filename.
args : typle
The arguments passed to the C function. Currently the code can
accept python ints and floats, as well as numpy numeric arrays
of all types. Numpy objects should always be after other
objects, and types should correspond with the definitions given
in py_types and np_types respectively.
py_types : typle of tuples (python_type, c_name)
Type specifications for non-numpy arguments. Currently only int
and float are valid types. Default is empty tuple.
np_types : typle of tuples (numpy_type, dims, c_name)
Type specifications for numpy-type arguments. Most numeric numpy
types are valid. dims is the integer number of dimensions of the
corresponding array. Default is empty tuple.
code : string, optional
C-code. One of code and code_path must be given.
code_path : string, optional
Full path to c-code. One of code or code_path should be given.
support_code : string, optional
C support code. This code will be inserted before the function
containing the c-code above. This can include any valid C code
including #includes and #defines.
support_code_path : string, optional
Full path to support code.
extension_kwargs : dictionary, optiona
Keyword arguments to pass to the distutils.Extension constructor.
return_type : python primitive type
Either int or float.
- inline_debug(unique_name, args=(), py_types=(), np_types=(), code=None, code_path=None, support_code=None, support_code_path=None, extension_kwargs={}, return_type=None)
- Same as inline, but the types of each argument are checked, and
the code is recompiled the first time this function is called.
|