Using Spark + Skylark to do large-scale spectral clustering with random feature maps - View it on GitHub