While parallel histograms and 2D graphs are good at emphasizing numerical relationships, grid maps and treemaps are useful visualizations to explore categorical and hierarchized relationships.

A simple grid is obtained by choosing to group (cluster) the cities by state, and displaying the groups on a grid. Notice that the view is recursive, and that inside each state, you can view all the cities belonging to that state, all layed out on a grid. This lets you view which states contain many cities. To ease the comparisons, you can sort the states by increasing population, and set the label of each state to show the total population held in the cities:

It is possible to navigate in this view, for instance to focus on one particular state to analyze the cities it contains. The present data set is too small to show more exciting findings, or to further partition the states in subcategories, but the sample data sets show numerous uses of grids and other maps applied to much larger data sets.