For the last couple of years, I’ve been itching to build a substantial project in Clojure.
Thanks to some new libraries, I think Clojure finally has the chance to compete on equal footing with Python. Mike Anderson’s
core.matrix offers a compelling alternative to
numpy and JFreeChart should be sufficient for my purposes. I’m most excited about Rinu Boney’s new Clojure machine-learning library
clatern, which has the potential to be a decent alternative to
scikit-learn for most of my day-to-day work.
Building and maintaining an open-source project is hard work. Building a community around it is even harder. Since
clatern scratches a real itch for me, I care about its success and decided to get involved. Rinu has already done the initial hard work of implementing a handful of useful methods and outlining a vision for the project.
To be successful, however, a project also needs some polish so that’s where I’ve been focusing my contributions. A testing approach needs to be devised to help developers makes changes with confidence. Contribution guidelines and common coding conventions need to be established. And a project needs good documentation and clear examples to help users get started easily and quickly.
Keep your eye on
clatern – I’m expecting to see some great things come out of this project.