Comparing between groups

Groups and clusters are usually defined using categorical or ordinal variables. For example, Species is a categorical variable grouping observation according to the species name in the Iris dataset. The assignments and cutree functions of Clustering return a vector with integers for encoding a categorical or ordinal variable. We consider that variable to be ordinal if the order of clusters has a meaning, for example, if they are related to the number of elements in the cluster. However, we will always visualize the assignation vector as a categorical variable when discriminating clusters.

We can use the position, color hue, and shape aesthetics to visualize and discriminate groups when using packages based on the Grammar of Graphics. Facets are also an excellent way to compare the same plot between groups. In Chapter 5, Introducing the Grammar of Graphics, we learned how to map categorical variables to those aesthetics and create facet plots using Gadfly...