Eigenvectors, generalized eigenvectors and stability

146 Views Asked by At

I'm currently studying the stability conditions for symplectic matrices, but I've some doubts that probably stem directly from basic linear algebra.

Let $M$ be a 4x4 real symplectic matrix.

Let $V$ be its eigenvector, with $e^{i\theta}$ as eigenvalue ($\theta \in \mathbb R$), i.e. $MV = e^{i \theta}V$. It is possibile to prove that in this case we have that $e^{-i \theta}$ is another eigenvalue with eigenvector $V^*$, i.e. $MV^* = e^{-i \theta}V^*$ (where the superscript $^*$ stands for "complex conjugate").

Suppose there is too a generalized eigenvector $W$, i.e. $MW = e^{i \theta}W + V$.

  • Does it mean that there's too a complex conjugate generalized eigenvector $W^*$, i.e. $MW^* = e^{-i \theta}W^* + V^*$?
  • How many invariant bidimensional subspaces does the matrix $M$ have?
  • Depending on the eigenvalues ($e^{\pm i \theta} = \pm 1$ or $e^{\pm i \theta} \neq \pm 1$), is there a difference in the number of those invariant subspaces?