Jacques Bertin's defined a relationship between data types and visual cues in Semiology of Graphics in 1967. A seminal set of guidelines for using position, size, hue, angle, orientation, brightness, texture and shape in the communication of categorical and continuous data.
In information visualisation dimensionality is one of the central issues. Visualising two ‘dimensions’ or variables is fairly simple, but what are our options for the third? Choosing how to map data attributes to visual attributes will determine how our plot looks, and how well it communicates it's message and our choices become more and more complicated as the number of dimensions increase (we'll just stick with three for now though).
‘All the interesting worlds (physical, biological, imaginary, human) that we seek to understand are inevitably and happily multivariate in nature.’ (Tufte 1990)
The top left chart communicates two variables only, the other three use colour, size and density to communicate their third data dimension.
This triangle-shaped ternary diagram or ‘simplex’ plot communicates three values for each data point by positioning them on three axes.
Alternatively we could use animation.
コメント