Field Theory / Linear Algebra

In a Field of Positive Characteristic, $A^p=I$ Does Not Imply that $A$ is Diagonalizable.

Problems and solutions in Linear Algebra

Problem 91

Show that the matrix $A=\begin{bmatrix}
1 & \alpha\\
0& 1
\end{bmatrix}$, where $\alpha$ is an element of a field $F$ of characteristic $p>0$ satisfies $A^p=I$ and the matrix is not diagonalizable over $F$ if $\alpha \neq 0$.
Add to solve later

Comment.

Remark that if $A$ is a square matrix over $\C$ with $A^k=I$, then $A$ is diagonalizable.
For a proof of this fact, see If a power of a matrix is the identity, then the matrix is diagonalizable

Thus, over a field of characteristic $p>0$ the condition $A^p=1$ dose not always imply that $A$ is diagonalizable.

Proof.

By induction, it is straightforward to see that
\[A^m=\begin{bmatrix}
1 & m\alpha\\
0& 1
\end{bmatrix}\] for any positive integer $m$.
Thus
\[A^p=\begin{bmatrix}
1 & p\alpha\\
0& 1
\end{bmatrix}=I\] since $p\alpha=0$ in the field $F$.

Since the eigenvalues of $A$ is $1$, if $A$ is diagonalizable, then there exists an invertible matrix $P$ such that $P^{-1}AP=I$, or $AP=P$.
Let $P=\begin{bmatrix}
a & b\\
c& d
\end{bmatrix}$. Then we have
\[AP=\begin{bmatrix}
a+\alpha c & b+\alpha d\\
c& d
\end{bmatrix}\] and this is equal to $P$, hence
\begin{align*}
a+\alpha c=a \text{ and } b+\alpha d=b \\
\end{align*}
Thus
\begin{align*}
\alpha c=0 \text{ and } \alpha d=0 \\
\end{align*}
If $\alpha \neq 0$, then $c=d=0$ but this implies that the matrix $P$ is non-invertible, a contradiction.
Therefore we must have $\alpha=0$.

Add to solve later

More from my site

Each Element in a Finite Field is the Sum of Two Squares Let $F$ be a finite field. Prove that each element in the field $F$ is the sum of two squares in $F$. Proof. Let $x$ be an element in $F$. We want to show that there exists $a, b\in F$ such that \[x=a^2+b^2.\] Since $F$ is a finite field, the characteristic $p$ of the field […]
How to Diagonalize a Matrix. Step by Step Explanation. In this post, we explain how to diagonalize a matrix if it is diagonalizable. As an example, we solve the following problem. Diagonalize the matrix \[A=\begin{bmatrix} 4 & -3 & -3 \\ 3 &-2 &-3 \\ -1 & 1 & 2 \end{bmatrix}\] by finding a nonsingular […]
Two Matrices with the Same Characteristic Polynomial. Diagonalize if Possible. Let \[A=\begin{bmatrix} 1 & 3 & 3 \\ -3 &-5 &-3 \\ 3 & 3 & 1 \end{bmatrix} \text{ and } B=\begin{bmatrix} 2 & 4 & 3 \\ -4 &-6 &-3 \\ 3 & 3 & 1 \end{bmatrix}.\] For this problem, you may use the fact that both matrices have the same characteristic […]
Determine Eigenvalues, Eigenvectors, Diagonalizable From a Partial Information of a Matrix Suppose the following information is known about a $3\times 3$ matrix $A$. \[A\begin{bmatrix} 1 \\ 2 \\ 1 \end{bmatrix}=6\begin{bmatrix} 1 \\ 2 \\ 1 \end{bmatrix}, \quad A\begin{bmatrix} 1 \\ -1 \\ 1 […]
Quiz 13 (Part 1) Diagonalize a Matrix Let \[A=\begin{bmatrix} 2 & -1 & -1 \\ -1 &2 &-1 \\ -1 & -1 & 2 \end{bmatrix}.\] Determine whether the matrix $A$ is diagonalizable. If it is diagonalizable, then diagonalize $A$. That is, find a nonsingular matrix $S$ and a diagonal matrix $D$ such that […]
Determine Dimensions of Eigenspaces From Characteristic Polynomial of Diagonalizable Matrix Let $A$ be an $n\times n$ matrix with the characteristic polynomial \[p(t)=t^3(t-1)^2(t-2)^5(t+2)^4.\] Assume that the matrix $A$ is diagonalizable. (a) Find the size of the matrix $A$. (b) Find the dimension of the eigenspace $E_2$ corresponding to the eigenvalue […]
Given Graphs of Characteristic Polynomial of Diagonalizable Matrices, Determine the Rank of Matrices Let $A, B, C$ are $2\times 2$ diagonalizable matrices. The graphs of characteristic polynomials of $A, B, C$ are shown below. The red graph is for $A$, the blue one for $B$, and the green one for $C$. From this information, determine the rank of the matrices $A, B,$ and […]
A Matrix Similar to a Diagonalizable Matrix is Also Diagonalizable Let $A, B$ be matrices. Show that if $A$ is diagonalizable and if $B$ is similar to $A$, then $B$ is diagonalizable. Definitions/Hint. Recall the relevant definitions. Two matrices $A$ and $B$ are similar if there exists a nonsingular (invertible) matrix $S$ such […]