ABSTRACT
In Chapter 3 we interpreted the positions of two-dimensional profile points in a triangular coordinate system where distances were Euclidean distances. In Chapter 4 the chi-square distance (χ2-distance) between profile points was defined, as well as its connection with the chi-square statistic and the inertia of a data matrix. The χ2-distance is a weighted Euclidean distance, where each squared term corresponding to a coordinate is weighted inversely by the average profile value corresponding to that coordinate. So far we have not actually visualized the χ2-distances between profiles, apart from Exhibit 4.2, where the average profile values were equal, so that the χ2-distances were also Euclidean in that case. In this chapter we show that by a simple transformation of the profile space, the distances that we observe in our graphical display are actual χ2-distances.
