Some colleagues and I are projecting population in North Dakota and we tried to figure out ways to identify different subregions of the state using k-means cluster analysis on a few economic variables.