In this tutorial we elaborate the Bauer Fike Theorem using an example of centered complex Gaussian matrix R perturbed by complex uniformly distributed matrix E, the theorem states that :

min |L-mu| < ||X|| ||X^-1|| ||E||

where (X,L) are the eigenvectors and eigenvalues of R respectively, mu is the spectrum of the perturbation.
The Frobenius norm is used in this simulation .

For more information, see this presentation from Eindhoven university :

http://www.win.tue.nl/casa/meetings/seminar/previous/_abstract051019_files/Presentation.pdf