Cython to speed up some code

[NardJ]

Stumbled accross Cython which may have some use to speed up PhotonFileEditor.

Nice comparison between Cython, CPython, PyPy and Python.

And here as simple benchmark.

[NardJ]

Nice demo of inline Cython:

https://stackoverflow.com/questions/11731428/finding-first-non-zero-value-along-axis-of-a-sorted-two-dimensional-numpy-array

[X3msnake]

Question is can we make it work? How would we go about. Does it need code rewriting No dia segunda-feira, 6 de agosto de 2018, Nard Janssens < [email protected]> escreveu:

…

Nice demo of inline Cython: https://stackoverflow.com/questions/11731428/finding- first-non-zero-value-along-axis-of-a-sorted-two-dimensional-numpy-array [image: image] <https://user-images.githubusercontent.com/11459480/43713325-816d62c0-9979-11e8-8b91-a2dba01a4a55.png> — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#80 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AKke-mH7YEBVohL881IGQBxvqV01Ifnoks5uOCNXgaJpZM4VqYdm> .

-- Com os melhores cumprimentos, Vinicius Silva

[NardJ]

Well I think a lot can be achieved with smarter coding and parallel processing. This should be a last resort because it would python collaborators to help with programming.

[NardJ]

Another demo calculating mandelbroth images.

[NardJ]

And from docs.cython.org
Building a Cython module using distutils¶

Imagine a simple “hello world” script in a file hello.pyx:

def say_hello_to(name):
    print("Hello %s!" % name)

The following could be a corresponding setup.py script:

from distutils.core import setup
from Cython.Build import cythonize

setup(name='Hello world app',
      ext_modules=cythonize("hello.pyx"))

To build, run python setup.py build_ext --inplace. Then simply start a Python session and do from hello import say_hello_to and use the imported function as you see fit.

One caveat if you use setuptools instead of distutils, the default action when running python setup.py install is to create a zipped egg file which will not work with cimport for pxd files when you try to use them from a dependent package. To prevent this, include zip_safe=False in the arguments to setup().

[NardJ]

Numba is very easy to implement, but not quite as fast as Cython...but very easy to implement

import numpy as np
from numba import autojit, jit

@autojit
def inc(arr, i, j):
    arr[i, j] += 1

@autojit
def test5(arr):
    for i in xrange(arr.shape[0]):
        for j in xrange(arr.shape[1]):
            inc(arr, i, j)

As the author on stackoverflow comments:

Even though it's 4.7x slower than Cython, most likely because the JIT compiler failed to inline inc(), I think it is AWESOME! All I needed to do is to add @autojit and didn't have to mess up the code with clumsy type declarations; 88x speedup for next to nothing!