Given that the jordan form of the matrix $A \in Mat_7(\mathbb F)$ is:

$\begin{pmatrix} J_2(1) &\cdots &0\\0& \cdots J_3(1) \cdots &0\\0&0& \cdots J_2(2)\end{pmatrix}$

Find the Jordan form of $A^2+A+I$

Clarification: $J_{\alpha}(\beta)$ is a jordan block of size $\alpha$ with $\beta$ on the diagonal.

What I did:

I tried bruteforcing it, as in actually calculating $A^2+A+I$, but now to find the minimal polynomial of a 7x7 matrix...not fun. Is there a smarter way to go about this question?

From the jordan form of $A$ I can infer the minimal polynomial and characteristic polynomial of $A$. Does that have any implication on the characteristic and minimal polynomials of $A^2+A+I$?

  • $\begingroup$ A tiny hint that may or may not be of specific use to you: $(A^2+A+I)(A-I)=A^3-I$... $\endgroup$ – Steven Stadnicki Feb 4 '14 at 19:28
  • $\begingroup$ I wonder if that's applicable here. $\endgroup$ – Oria Gruber Feb 4 '14 at 19:39

Hint: If $$ A = SJS^{-1} $$ then

$$A^2+A+I = \left(SJS^{-1}\right)^2+\left(SJS^{-1}\right)+SS^{-1} = S\left(J^2+J+I\right)S^{-1}$$

  • $\begingroup$ Brilliant!!!!...But that doesn't help that much. Now I need to find the Jordan form of $J^2+J+I$, thats still a 7 by 7 matrix $\endgroup$ – Oria Gruber Feb 4 '14 at 19:29
  • $\begingroup$ $(J^2+J+I)$ has the same amount of "blocks" as $J$ and you only have to handle each of those blocks independently. Those blocks are either Jordan blocks, or triangular and easy to get into JNF $\endgroup$ – user127.0.0.1 Feb 4 '14 at 19:47
  • $\begingroup$ So I don't need to actually find the Jordan form of $J^2+J+I$? I can find the jordan form of each block individually and infer from that the jordan form of the matrix im looking for? $\endgroup$ – Oria Gruber Feb 4 '14 at 19:50
  • $\begingroup$ Yes, that's right. $\endgroup$ – user127.0.0.1 Feb 4 '14 at 20:28

We have $C_{A}(x) = (x-1)^{5}(x-2)^{2}$ and $m_{A}(x) = (x-1)^{3}(x-2)^{2}$. Write $B = A^{2} + A + I$. Then $Av = \lambda v \implies Bv = (\lambda^{2} + \lambda + 1)v$. This equation tells us what the eigenvalues of $B$ look like, but also more: $ \ker(A-\lambda I) \subset \ker \left (B - (\lambda^{2} + \lambda + 1)I\right)$ and the reverse inclusion also holds. Therefore, the structure of $J_{A}$ is preserved. Thus, $C_{B}(x) = (x-3)^{5}(x-7)^{2}$ and $m_{B}(x) = (x-3)^{3}(x-7)^{2}$.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.