Functions for detecting matrices with Jordan blocks

Detecting a nearest matrix (or values of matrix parameters), when the matrix has a Jordan block.

This collection of functions assumes a real or complex matrix or a multi-parameter matrix family as an input. It provides the nearest matrix or values of parameters, when a Jordan block of a given order appears and computes the correponding eigenvalue and generalized eigenvectors.

The numerical method is based on the versal deformation theory and is described in the paper: Alexei A. Mailybaev, Computation of multiple eigenvalues and generalized eigenvectors for matrices dependent on parameters, Numer. Linear Algebra Appl. 13(2006), 419–436.

There is a description in the readme.pdf file and some specifications in the file of each function.