Diagonalizable Matrix with Eigenvalue 1, -1

Stanford University Linear Algebra Exam Problems and Solutions

Problem 37

Suppose that $A$ is a diagonalizable $n\times n$ matrix and has only $1$ and $-1$ as eigenvalues.
Show that $A^2=I_n$, where $I_n$ is the $n\times n$ identity matrix.

(Stanford University Linear Algebra Exam)

See below for a generalized problem.

LoadingAdd to solve later

Sponsored Links


Hint.

Diagonalize the matrix $A$ so that the diagonal matrix has $\pm 1$ on diagonal entries.

Proof.

Since $A$ is diagonalizable, there exists an invertible matrix $P$ such that $P^{-1}AP=D$, where $D$ is a diagonal matrix.
Since $A$ has only $\pm 1$ as eigenvalues, we can choose $P$ so that the diagonal entries of $D$ are either $\pm 1$.

Then we have $A=PDP^{-1}$ and \(\require{cancel}\)
\begin{align*}
A^2=(PDP^{-1})(PDP^{-1})=(PD\cancel{P}^{-1})(\cancel{P}DP^{-1})=PD^2P^{-1}.
\end{align*}
Note that $D^{2}=I_2$ and thus we have $A^2=I_2$ as required.

Generalization

As a generalization, consider the following problem.

Problem.
Suppose that $A$ is a diagonalizable $n\times n$ matrix and the eigenvalues of $A$ are $r$-th roots of unity for some positive integer $r$.
Show that $A^r=I_n$, where $I_n$ is the $n\times n$ identity matrix.

The proof of Problem 2 is a straightforward generalization of the proof of Problem 1. So give it a shot.

Related Problem.

For a similar problem, see the post ↴
 Finite order matrix and its trace.

This problem is the converse of the current problem in some sense.


LoadingAdd to solve later

Sponsored Links

More from my site

You may also like...

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

More in Linear Algebra
Purdue University Linear Algebra Exam Problems and Solutions
Find a Formula for a Linear Transformation

If $L:\R^2 \to \R^3$ is a linear transformation such that \begin{align*} L\left( \begin{bmatrix} 1 \\ 0 \end{bmatrix}\right) =\begin{bmatrix} 1 \\...

Close