Provides several classes for computing various linear components of sets of vectors.
For instance, mcal_pca performs Principle Component Analysis
The mcal_general_ca generates a set of orthogonal axes so as to minimise some cost function based on the axes and the projection onto them, encoded in classes derived from mcal_single_basis_cost.
One can perform a version of Sparse PCA by plugging a mcal_sparse_basis_cost object into the mcal_general_ca.