@mbostock Pushed 1d k-means clustering into a d3-fork, but have no idea how to create a d3-scale. Too confusing.. http://t.co/hCvbtPZr