Skip to main content

Section 31.4 Examples

Subsection 31.4.1 Matrix polynomials

Example 31.4.1. Evaluating a matrix polynomial.

Consider again the polynomial

\begin{equation*} p(x) = x^2 - 3x + 6 \end{equation*}

from Discovery 31.1. As in Subsection 31.3.1, we rewrite this as

\begin{equation*} p(X) = X^2 - 3X + 6I_n \text{,} \end{equation*}

where the indeterminate \(X\) now represents an \(n \times n\) matrix. We have also written \(I_n\) to represent the \(n \times n\) identity matrix so that we can be specific about the size of identity used.

First, let's try substituting the \(2 \times 2\) matrix

\begin{equation*} A = \left[\begin{array}{rc} 1 \amp 2 \\ -1 \amp 3 \end{array}\right] \end{equation*}

into \(p(X)\text{:}\)

\begin{align*} p(A) \amp= A^2 - 3A + 6I_2 \\ \amp= \left[\begin{array}{rc} -1 \amp 8 \\ -4 \amp 7 \end{array}\right] - \left[\begin{array}{rc} 3 \amp 6 \\ -3 \amp 9 \end{array}\right] + \begin{bmatrix} 6 \amp 0\\ 0 \amp 6 \end{bmatrix}\\ \amp= \left[\begin{array}{rc} 2 \amp 2 \\ -1 \amp 4 \end{array}\right] \text{.} \end{align*}

Now here's an example of substituting the diagonal \(3 \times 3\) matrix

\begin{equation*} D = \left[\begin{array}{crc} 5 \amp 0 \amp 0 \\ 0 \amp -1 \amp 0 \\ 0 \amp 0 \amp 4 \end{array}\right] \end{equation*}

into \(p(X)\text{.}\) Notice how the result matrix is diagonal with the result of evaluating the polynomial at each diagonal entry down the main diagonal.

\begin{align*} p(D) \amp= D^2 - 3D + 6I_3 \\ \amp = \begin{bmatrix} 25 \amp 0 \amp 0\\ 0 \amp 1 \amp 0\\ 0 \amp 0 \amp 16 \end{bmatrix} - \left[\begin{array}{crc} 15 \amp 0 \amp 0\\ 0 \amp -3 \amp 0\\ 0 \amp 0 \amp 12 \end{array}\right] + \begin{bmatrix} 6 \amp 0 \amp 0\\ 0 \amp 6 \amp 0\\ 0 \amp 0 \amp 6 \end{bmatrix}\\ \amp= \begin{bmatrix} 16 \amp 0 \amp 0\\ 0 \amp 10 \amp 0\\ 0 \amp 0 \amp 10 \end{bmatrix} \end{align*}

Subsection 31.4.2 Nilpotent matrices

Example 31.4.2. Patterns of super- and sub-diagonals in powers of a triangular nilpotent matrix.

Here are two examples, one upper and one lower triangular, of powers of nilpotent matrices. Notice how the main diagonal of zeros marches up the super-diagonals of the upper triangular powers and down the sub-diagonals in the lower triangular powers, until eventually the entire matrix becomes zero. (We have omitted the zeros below the main diagonal in the upper triangular example and above the main diagonal in the lower triangular example to emphasize this pattern.)

\begin{align*} U \amp = \begin{bmatrix} 0 \amp 1 \amp 1 \amp 1 \amp 1 \\ \amp 0 \amp 1 \amp 1 \amp 1 \\ \amp \amp 0 \amp 1 \amp 1 \\ \amp \amp \amp 0 \amp 1 \\ \amp \amp \amp \amp 0 \end{bmatrix} \amp L \amp = \begin{bmatrix} 0 \\ 1 \amp 0 \\ 1 \amp 1 \amp 0 \\ 1 \amp 1 \amp 1 \amp 0 \\ 1 \amp 1 \amp 1 \amp 1 \amp 0 \end{bmatrix}\\ U^2 \amp = \begin{bmatrix} 0 \amp 0 \amp 1 \amp 2 \amp 3 \\ \amp 0 \amp 0 \amp 1 \amp 2 \\ \amp \amp 0 \amp 0 \amp 1 \\ \amp \amp \amp 0 \amp 0 \\ \amp \amp \amp \amp 0 \end{bmatrix} \amp L^2 \amp = \begin{bmatrix} 0 \\ 0 \amp 0 \\ 1 \amp 0 \amp 0 \\ 2 \amp 1 \amp 0 \amp 0 \\ 3 \amp 2 \amp 1 \amp 0 \amp 0 \end{bmatrix}\\ U^3 \amp = \begin{bmatrix} 0 \amp 0 \amp 0 \amp 1 \amp 3 \\ \amp 0 \amp 0 \amp 0 \amp 1 \\ \amp \amp 0 \amp 0 \amp 0 \\ \amp \amp \amp 0 \amp 0 \\ \amp \amp \amp \amp 0 \end{bmatrix} \amp L^3 \amp = \begin{bmatrix} 0 \\ 0 \amp 0 \\ 0 \amp 0 \amp 0 \\ 1 \amp 0 \amp 0 \amp 0 \\ 3 \amp 1 \amp 0 \amp 0 \amp 0 \end{bmatrix}\\ U^4 \amp = \begin{bmatrix} 0 \amp 0 \amp 0 \amp 0 \amp 1 \\ \amp 0 \amp 0 \amp 0 \amp 0 \\ \amp \amp 0 \amp 0 \amp 0 \\ \amp \amp \amp 0 \amp 0 \\ \amp \amp \amp \amp 0 \end{bmatrix} \amp L^4 \amp = \begin{bmatrix} 0 \\ 0 \amp 0 \\ 0 \amp 0 \amp 0 \\ 0 \amp 0 \amp 0 \amp 0 \\ 1 \amp 0 \amp 0 \amp 0 \amp 0 \end{bmatrix}\\ U^5 \amp = \zerovec \amp L^5 \amp = \zerovec \end{align*}
Example 31.4.3. Scalar-triangular matrices with \(\lambda = 0\).

A scalar-triangular matrix with \(\lambda = 0\) repeated down the main diagonal is always nilpotent. The \(5 \times 5\) matrices in Example 31.4.2 were both of this form, but it is not necessary to have a super- or sub-diagonal of ones to be nilpotent.

\begin{align*} U_2 \amp= \begin{bmatrix} 0 \amp a \\ 0 \amp 0 \end{bmatrix} \amp U_3 \amp= \begin{bmatrix} 0 \amp a \amp b \\ 0 \amp 0 \amp c \\ 0 \amp 0 \amp 0 \end{bmatrix} \amp U_4 \amp= \begin{bmatrix} 0 \amp a \amp b \amp c \\ 0 \amp 0 \amp d \amp e \\ 0 \amp 0 \amp 0 \amp f \\ 0 \amp 0 \amp 0 \amp 0 \end{bmatrix}\\ U_2^2 \amp= \zerovec \amp U_3^2 \amp= \begin{bmatrix} 0 \amp 0 \amp ac \\ 0 \amp 0 \amp 0 \\ 0 \amp 0 \amp 0 \end{bmatrix} \amp U_4^2 \amp= \begin{bmatrix} 0 \amp 0 \amp bd \amp ae + bf \\ 0 \amp 0 \amp 0 \amp df \\ 0 \amp 0 \amp 0 \amp 0 \\ 0 \amp 0 \amp 0 \amp 0 \end{bmatrix}\\ \amp \amp U_3^3 \amp= \zerovec \amp U_4^3 \amp= \begin{bmatrix} 0 \amp 0 \amp 0 \amp adf \\ 0 \amp 0 \amp 0 \amp 0 \\ 0 \amp 0 \amp 0 \amp 0 \\ 0 \amp 0 \amp 0 \amp 0 \end{bmatrix} \amp\\ \amp \amp \amp \amp U_4^4 \amp= \zerovec \end{align*}
Example 31.4.4. A non-triangular nilpotent matrix.

Matrices can be nilpotent without being scalar-triangular. And while we will see that the size of a nilpotent matrix is the “latest” that its powers become zero (Theorem 31.5.3 in Subsection 31.5.1), it can occur before that too.

\begin{align*} N \amp = \left[\begin{array}{rrrrr} 1 \amp -2 \amp -1 \amp -1 \amp 0 \\ -1 \amp -2 \amp -2 \amp 0 \amp 1 \\ 0 \amp 1 \amp 1 \amp 0 \amp 0 \\ 3 \amp 1 \amp 2 \amp -1 \amp -2 \\ 0 \amp -3 \amp -2 \amp -1 \amp 1 \end{array}\right]\\ N^2 \amp = \left[\begin{array}{rrrrr} 0 \amp 0 \amp 0 \amp 0 \amp 0 \\ 1 \amp 1 \amp 1 \amp 0 \amp -1 \\ -1 \amp -1 \amp -1 \amp 0 \amp 1 \\ -1 \amp -1 \amp -1 \amp 0 \amp 1 \\ 0 \amp 0 \amp 0 \amp 0 \amp 0 \end{array}\right]\\ N^3 \amp = \zerovec \end{align*}

In Section 33.5, we will see that this particular nilpotent \(5 \times 5\) matrix becomes zero at the third power instead of just at the fifth power because it has a smaller \(3 \times 3\) nilpotent block “hiding” inside it. (In particular, see Example 33.5.4.)