@mjs2600 I may be biased but I think core.matrix is fundamentally better than Numpy in terms of design. And can easily be faster if you use a decent implementation. Not quite all the features yet, but easy enough to add
Hi all I am experimenting with a new native matrix library for Clojure, building on netlib-java. Would appreciate comments / thoughts. https://github.com/mikera/vectorz-native
Basically it is a "plugin" library for vectorz-clj that uses BLAS
It looks awesome! I'll try it for my RL algorithms. I'll let you know how it goes.