Cluster structures by fingerprints and inspect the clustering statistics to choose a good number of clusters. Create automatically the optimum number of clusters based on the Kelley penalty. Select diverse representatives per cluster. [Requires: Canvas]