A SPHERICAL REPRESENTATION OF A CORRELATION MATRIX

It is common practice to perform a principal component analysis (PCA) on a correlation matrix to represent graphically the relations among n umerous variables. In such a situation, the variables may be considere d as points on the unit hypersphere of an Euclidean space, and PCA pro vides a sort of best fit of these points within a subspace. Taking int o account their particular position, this paper suggests to represent the variables on an optimal three-dimensional unit sphere.