fnands · 87 points · 33 comments · Oct 6 · Open original
Comments
5 preview comments · loading full thread
Log in to use comments
Log in to h4cker, then connect Hacker News to publish comments.
HYhyperbovineOct 7
In my niche corner of scientific computing it feels like Cython has largely been replaced by Numba and CFFI, or just Julia. Last I checked it still needed setup.py which is a bit of a deal breaker in 2025.
ADadsharmaOct 6
A more interesting path is to keep dbscan_inner in pure python with type annotations and then use
py2many --mojo=1 dbscan_inner.py
to translate.
RORossBencinaOct 7
Very interesting. I'm currently trading off whether to use Mojo or C++/pybind to accelerate simulations that combine matrix operations with fine-grained scalar calculations. I only recently learned that pybind + cppimport offers the integrated compile-on-import experience available in Mojo.
PJpjmlpOct 7
> I think moving a lot of scikit-learn’s more computationally intensive code to Mojo could be an interesting project.
Only if you want to lose access to Windows users, as it is a low priority for Mojo development.
JOjononorOct 8
Mojo is not open source, so how can it be realistic to use it in scikit-learn?
We spent decades getting out of the clutches of Mathworks, Microsoft, etc. Why are people eager to go back that way?
Comments
5 preview comments · loading full threadLog in to h4cker, then connect Hacker News to publish comments.
In my niche corner of scientific computing it feels like Cython has largely been replaced by Numba and CFFI, or just Julia. Last I checked it still needed setup.py which is a bit of a deal breaker in 2025.
A more interesting path is to keep dbscan_inner in pure python with type annotations and then use py2many --mojo=1 dbscan_inner.py to translate.
Very interesting. I'm currently trading off whether to use Mojo or C++/pybind to accelerate simulations that combine matrix operations with fine-grained scalar calculations. I only recently learned that pybind + cppimport offers the integrated compile-on-import experience available in Mojo.
> I think moving a lot of scikit-learn’s more computationally intensive code to Mojo could be an interesting project. Only if you want to lose access to Windows users, as it is a low priority for Mojo development.
Mojo is not open source, so how can it be realistic to use it in scikit-learn? We spent decades getting out of the clutches of Mathworks, Microsoft, etc. Why are people eager to go back that way?