Hi! For a set of rows like on the picture below, is there a fully automated classification method (i.e. requiring no configuration) that would be able to split the set in two disjoint subsets? I, for one, clearly see two sets, namely, more dense "on the left" and the opposite of that one. A Clojure lib would be a plus, of course. :)
It's rather a task for clustering. And there is a library for that in Clojure. Check out the suite for algorithms. But note that ML always requires some tuning and data preparation.
As I was fiddling with it I realized that there is rather a continuum to it. Upon applying a certain order, the full picture looks like below. I will definitely check out scicloj, thank you for the pointer!