Skip to content

6.2-exam3-1998-strang

Exam 3, 1998 (strang) · Section 6.2: Diagonalizing a Matrix Source: exam3_1.pdf, pp. 25, 28

Problem

2 (a) (10 points) Find the eigenvalues \(\lambda_1, \lambda_2, \lambda_3\) (NOT the eigenvectors \(x_1, x_2, x_3\)) of this Markov matrix:

\[ A = \begin{bmatrix} .6 & .6 & 0 \\ .2 & .2 & .2 \\ .2 & .2 & .8 \end{bmatrix} \]

(b) (10 points) Suppose \(u_0\) is the sum \(x_1 + x_2 + x_3\) of the three eigenvectors that you didn't compute. What is \(A^n u_0\)?

(c) (4 points) As \(n \to \infty\) what is the limit of \(A^n u_0\)?

Solution (a) \[ \lambda_1 = 1 \quad (\text{for any Markov matrix}), \] \[ \lambda_2 = 0 \quad (\text{since } A \text{ is singular}), \] \[ \lambda_3 = .6 \quad (\text{since the trace of } A \text{ is } 1.6). \] (1, 2, 3 can be permuted.) (b) For \(n \geq 1\): \[ A^n u_0 = 1^n x_1 + 0^n x_2 + (.6)^n x_3 = x_1 + (.6)^n x_3. \] (c) \(A^n u_0\) approaches \(x_1\).

Keywords: Markov matrix, eigenvalues, matrix powers, steady state