KMeans Clustering


A few years ago I read a book on artificial intelligence algorithms (Artificial Intelligence for Humans) and had it on my todo list to implement a couple of the algorithms.

There’s lots of implementations out there already in Go, but I’ve always found that writing an implementation makes sure I understand how something really works and why existing implementations are designed the way they are, or what problems to look out for.

So far, I’ve implemented KMeans clustering and a few distance metrics to use: Chebyshev (Chessboard), Euclidean and Manhattan.

Here’s some example code which produces the PNG graph above.

The code is up at

I’ll be adding more functions when I feel like it.