Profiling your data using parallel coordinate plots

The Parallel Coordinate Plot visualization is used to compare a set of potentially diverse and unrelated properties that can nevertheless be attached to a themed series. A typical example would be comparing the specifications of a selection of desktop computers. The properties can be anything from keyboard color to processor speed, but they all apply to each PC in our selection.

The columns in a parallel coordinate plot are the properties we want to include, and their values, whether numbers or text, are normalized based on the value for numbers and an inferred value for strings based on natural string ordering. This normalization is the key to a parallel coordinate plot because it allows us to compare quantitative and qualitative information in the same plot. For example, if you include keyboard color and price in a comparison of 10 PCs, the color furthest down the sort order would be assigned the value of 100 percent, as would the highest...