# Are the following equations concerning matrices and their eigenvectors equivalent?

Let $A$ be a matrix, $a,b$ its eigenvectors and $\lambda_a, \lambda_b$ their corresponding eigenvalues. Define $a^\prime = Aa$ and $b^\prime = Ab$. Is it true that $$\lambda_a = \frac{\Vert a^\prime\Vert}{\Vert a\Vert} \iff \Vert a\Vert = \frac{1}{\sqrt{\lambda_a}}$$ and $$\lambda_b = \frac{\Vert b^\prime\Vert}{\Vert b\Vert} \iff \Vert b\Vert = \frac{1}{\sqrt{\lambda_b}} ?$$

No. First of all, note that the statements don't even make sense if $\lambda_a=0$ or $\lambda_b=0$. On the other hand, $a'=\lambda_aa$ and therefore$$\frac{\|a'\|}{\|a\|}=\lvert\lambda_a\rvert.$$So, $\lambda_a=\frac{\|a'\|}{\|a\|}\iff\lambda_a\geqslant0$. And, of course, $\lambda_a\geqslant0$ is not equivalent to $\|a\|=\frac1{\sqrt{\lambda_a}}$.
Hint: $$\frac{||a'||}{||a||}=\frac{||Aa||}{||a||}=\frac{||\lambda_aa||}{||a||}=\frac{|\lambda_a|\cdot||a||}{||a||}=|\lambda_a|$$
Hint : $$a' = A a = \lambda_a a$$