I have the following orthogonal matrix: $M_{5,5}=\left( \begin{array}{ccccc} 0 & 1 & 0 & 0 & 0 \\ 1 & 0 & 0 & 0 & 0 \\ 0 & 0 & 0 & 1 & 0 \\ 0 & 0 & 0 & 0 & 1 \\ 0 & 0 & 1 & 0 & 0 \\ \end{array} \right)$ My aim here is to diagonalize it and check some interesting properties of powers of $M$. Then I proceed to express $M$ as $M=P.D.P^{-1}$ >$\{d,P\}=\text{Eigensystem}[M]$ >$\text{P = Transpose[P]}$ >$\text{d = DiagonalMatrix[d]}$ >$\text{Round}\left[P.d.P^{-1}\right]==M$ which is **True** >$\text{Round}\left[P.d.d.P^{-1}\right]==M.M$ which is **True** also. This means that $M^k = P.D^k.P^{-1}$ so the eigenvalues of $A^k$ are $\lambda^k=\{D^k_{1,1},D^k_{2,2},D^k_{3,3},D^k_{4,4},D^k_{5,5}\}$ but when I compare $\lambda^2$ with the eigenvalues of $M^2$ the output is **False**. >$\text{DiagonalMatrix}[\text{Eigensystem}[M.M][[1]]]=d.d$ which is **False** Now I start to think that $M^2$ can be written in 2 different ways: $M^2 = P.D^2.D^{-1}$ which we have seen it is true. $M^2 = Q.D'.Q^{-1}$ this means that other eigen decomposition exists in ${M^2}$. Can someone tell me If I'm wrong? If true why $M^2$ can be written in these two expressions?