Maps and Hierarchies

Maps of various kinds (grids, treemaps, and others) are useful when your data is hierarchized, or when you want to see how the data is distributed according to one or more dominant criteria:

Grids enable you to view data split into multiple categories, each category being shown as a small square whose representation can be customized to display relevant information on this category.
Hierarchized views. Grids and other maps can be embedded in one another, showing the relative distribution of a subcategory according to a main category.
Treemaps (see http://www.cs.umd.edu/hcil/treemap-history) are one kind of hierarchized view that can be produced with ILOG Discovery.

Grids

We have grouped a database of US cities by state, and laid out those states in a grid. For each state, we show a 2D plot of Housing Cost vs. Climate, enabling between-states comparisons of the correlation of both factors.

Hierarchized Views

Here, we show some insurance premium data for a population of drivers. The drivers are grouped according to the type of Airbag they have in their car, and their vehicle type. The colors indicate the number of accidents each driver has had in the past.

This view allows assessing whether different vehicle types and airbag types are correlated with the number of accident or the average premium paid by this driver.

Treemaps

Here we show a hierarchy of 50000 files, colored according to their file type.