Homework 5

Due Date:

Wednesday, Week 7

Textbook Chapters:

5.1,5.2,5.3

1

By hand, find eigensystems for each of these matrices. Then specify the algebraic and geometric multiplicity of each eigenvalue.

  1. [71058]

  2. [100110011]

  3. [211031002]

  4. [0220]

  5. [0220]

2

Compute the eigensystems of these matrices, and identify any defective matrices. (You can do this by hand or on the computer.)

  1. [201000102]

  2. [200031062]

  3. [1+i30i]

  4. [121242001]

  5. [2112011200010010]

3

By hand, find a matrix P such that P1AP is diagonal for each of the following matrices. Then use P to find a formula for Ak,k a positive integer.

  1. [201010003]

  2. [122000022]

  3. [1232]

  4. [0220]

  5. [2100001000310001]

4

Compute sin(π6A) and cos(π6A), where A=[2403].

5

By hand, find the spectral radius and dominant eigenvalue, if any, for each of these matrices:

  1. [7698]

  2. 13[1320]

  3. [010001100]

  4. 12[101100021]

  5. [1111]

6

If the matrices of the previous exercise are transition matrices, for which do all x(k) remain bounded as k ? Are any of these matrices stable?

7

The three-stage insect model of Example 2.21 yields a transition matrix

A=[0.200.250.60.3000.60.8]

Use a technology tool to calculate the eigenvalues of this matrix. Deduce that A is diagonalizable and determine the approximate growth rate from one state to the next (after much time has passed), given a random initial vector.

8

The financial model of Example 2.27 gives rise to a discrete dynamical system x(k+1)=Ax(k), where the transition matrix is

A=[10.060.12100010]

Use a technology tool to calculate the eigenvalues of this matrix. Deduce that A is diagonalizable and determine the approximate growth rate from one state to the next, given a random initial vector.

Find a starting vector x(0) such that growth rate for the first few iterations is quite different from this approximate growth rate, and demonstrate this using a technology tool.

(If you do this using Sympy, and you set up the matrix A using rational numbers, you can get exact results and find a starting vector which will never reach your approximate growth rage. You can see how this result changes if you use floating point numbers at some point in the calculation, e.g. x = N(x).)

from sympy import Matrix, Rational

# Set up the matrix A in symbolic form
A = Matrix([[1, Rational(6, 100), Rational(12, 100)], [1, 0, 0], [0, 1, 0]])

9

On this problem,you may use a computer to find eigensystems etc, but don’t solve it using simulations and trial and error.

A species of bird can be divided into three age groups: age less than 2 years for group 1, age between 2 and 4 years for group 2, and age between 4 and 6 years for the third group. Assume that these birds have at most a 6-year life span. It is estimated that the survival rates for birds in groups 1 and 2 are 50% and 75%, respectively. Also, birds in groups 1, 2, and 3 produce 0,1 , and 3 offspring on average in any biennium (period of 2 years).

  1. Model this bird population as a discrete dynamical system and analyze the long-term change in the population – give a percentage by which the population of group 1 changes from year to year.

  2. Now suppose that the survival rates for group 1 and 2 are not known (so are no longer assumed to be 50% and 75%), but are assumed to be equal to one another. What value for this survival rate would make the population stable in the long run?