Gower, John and Harding, Simon A.
(1998). Prediction regions for categorical variables.
In: Blasius, Jörg and Greenacre, Michael eds.
Visualization of Categorical Data.
Elsevier Inc., pp. 405–419.
Quantitative information on two or more variables is often represented relative to coordinate axes. In this chapter we show how the familiar concepts associated with quantitative axes may be extended to categorical variables. We shall be concerned with low-dimensional approximations to high-dimensional representations relative to coordinate axes, and to understand the properties of the approximation we must first recapitulate the familiar properties of Cartesian coordinate axes. Figure 1 shows two coordinate axes referring to variables x1 and x2, marked with scales
Actions (login may be required)