Skip to content

6.2-exam3-2008

Quiz 3, 2008 · Section 6.2: Diagonalizing a Matrix Source: exam3_1.pdf, pp. 139, 142

Problem

4 (25 pts.) In terms of \(x\) (\(0 < x < 1\)) complete

\[ A = \begin{bmatrix} x & \\ & \end{bmatrix}, \]

so that \(A\) is a 2x2 matrix that is both Markov and singular.

What is \(A^{2008}\)?

Solution As \(A\) is supposed to be Markov, we must have \(A = \begin{bmatrix} x & y \\ 1 - x & 1 - y \end{bmatrix}\), and \(A\) singular implies \[ x(1 - y) - y(1 - x) = 0, \quad \text{therefore} \quad 0 = x - xy - y + xy = x - y, \] so \(y = x\). Thus \[ A = \begin{bmatrix} x & x \\ 1 - x & 1 - x \end{bmatrix}. \] As \(A\) is Markov, we know that \(\lambda_1 = 1\) is an eigenvalue. As \(A\) is singular, we know that the product of the eigenvalues is 0. Therefore, \(\lambda_2 = 0\) is another eigenvalue, and so \(A\) is diagonalizable: \[ A = S \begin{bmatrix} 1 & 0 \\ 0 & 0 \end{bmatrix} S^{-1}. \] Thus \[ A^{2008} = S \begin{bmatrix} 1 & 0 \\ 0 & 0 \end{bmatrix}^{2008} S^{-1} = S \begin{bmatrix} 1 & 0 \\ 0 & 0 \end{bmatrix} S^{-1} = A. \]

Keywords: Markov matrix, singular matrix, diagonalizable, matrix powers