data-science

Data science, data analysis, and machine learning in Clojure https://scicloj.github.io/pages/chat_streams/ for additional discussions
viesti 2019-12-01T08:46:03.193800Z

Yup, just happened to have a moment of time to experiment :)

🦜 1
chrisn 2019-12-01T18:59:39.195700Z

@viesti - Also, you can do all of that via kind of stuff via libpython-clj: https://github.com/cnuernber/libpython-clj/tree/master/example/src

chrisn 2019-12-01T19:03:11.197500Z

Actually and while we are talking about it javacpp has an extensive set of bindings to many different things (all done way early): https://github.com/bytedeco/javacpp-presets I would much rather have JNA bindings than javacpp bindings but the javacpp bindings are there and are on maven in a format that allows them to just work with multiple operating systems. They precompile native deps and use the maven selectors to select the deps for your OS.

viesti 2019-12-17T10:34:18.009600Z

ah true, had not thought about looking for other native bindings

viesti 2019-12-17T10:34:33.009800Z

but yeah, lots of options 🙂

chrisn 2019-12-01T19:04:21.198100Z

Lots of different options 🙂.

2019-12-01T22:22:28.200Z

How stable do you think the API is on libpython-clj?

chrisn 2019-12-02T14:14:34.205900Z

It is pretty stable. The basic import-module, call attribute pathway hasn't changed since I announced it and that is really the only true pathway. Everything else on top of that is sugar to some extent or another.