0
$\begingroup$

We have a positive integer $n$ and two $n\times n$ matrices of real numbers, $A_{1}$ and $A_{2}$. For $j=1, 2$, we have the eigenvalues and eigenvectors $\lambda _{j}$ and $x_{j}$ of $A_{j}$. Show that, if $ A_{2}=A_{1}^{T}$, then $\lambda _{1}= \lambda _{2}$ or $x_{1}, x_{2}$ are vertical.

We have that $det(A_{1}-\lambda I_{n})=det(A_{2}-\lambda I_{n})$. Am I right? What can I do to solve this problem? I am really stuck.

$\endgroup$
0
$\begingroup$

Since the determinants are the same, you obtain the same characteristic polynomial for both matrices and therefore yield identical eigenvalues to both.

$\endgroup$
0
$\begingroup$

Let $A_1x_1 = \lambda_1x_1$ and $A_2x_2 = \lambda_2x_2$. Then $$ \lambda_1x_2^Tx_1 = x_2^T(\lambda_1x_1) = x_2^T(A_1x_1) = x_2^TA_1x_1 = (x_2^TA_1)x_1 = (x_2^TA_2^T)x_1 = \lambda_2x_2^Tx_1 $$ which in turn gives us $$ (\lambda_1-\lambda_2)(x_2^Tx_1) = 0 $$ which is what we wanted to show.

$\endgroup$
0
$\begingroup$

Hint

For any matriz $C_{n\times n}$ we know that $\det C=\det C^T$. Now, see that

$$(A_1-\lambda I)^T=A_1^T-\lambda I$$

Can you finish?

$\endgroup$

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.