Whenever using lda(...,CV=TRUE), or if using lda.predict(), a classification is returned. Which distance is used in order to allocate a given individual to a class, once the original data have been transformed into discriminant components? Thanks Agustin Lobo aloboaleu using gmail.com