ColorViz is an interactive color visualization tool that I created for my Intro to Info Viz final project at Georgia Tech. We were given full creative freedom in choosing what data to visualize, and I found myself drawn towards the color data of arbitrary images. K-means clustering allows a user to "clean" this data (using the ml-kmeans library loaded via CDN), transforming complex images into interpretable color palettes. It is also just a nifty way to understand what k-means does through a playful interaction 🙂 I use D3's heatmap and chord diagram modules to display this data.
