Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Repeated Eigenvalues

  1. Apr 12, 2008 #1
    1. The problem statement, all variables and given/known data

    Obtain the eigenvalues and corrosponding eigenvectors for the matrix: [2,2,1;1,3,1;1,2,2]

    2. Relevant equations

    3. The attempt at a solution

    I can solve for the eigenvalues, 5, 1, and 1

    I can solve the eigenvalue 5 for the eigenvector B[1;1;1]

    Yet somehow, the 1 eigenvalues are supposed to give me two different eigenvectors. Unfortunately, this makes no sense to me, which is why i'm here...

    Can someone please explain to me how to finish solving this question?
  2. jcsd
  3. Apr 12, 2008 #2


    User Avatar
    Science Advisor

    Since 1 is an eigenvaue, the equation
    [tex]\left[\begin{array}{ccc}2 & 2 & 1\\1 & 3 & 1\\ 1 & 2 & 2\end{array}\right]\left[\begin{array}{c} x \\ y \\ z\end{array}\right]= \left[\begin{array}{c}x \\ y \\ z\end{array}\right][/tex]
    must have a non-trivial solution. Multiplying that out and looking at individual rows, we must have 2x+ 2y+ z= x, x+ 3y+ z= y, x+ 2y+ 2z= z. Because 1 is an eigenvalue, those are not independent and must have an infinite number of solutions. In fact, in this particular problem, the three equations those reduce to, x+ 2y+ z= 0, x+ 2y+ z= 0, and x+ 2y+ z= 0, are all the same equation. You can solve for z as a function of x and y: z= -x- 2y. That means you can choose any numbers you like for x and y, then solve for z to get an eigenvalue. Can you choose them so that you get two independent vectors? (I always like to use simple numbers like 0 and 1.)
  4. Apr 12, 2008 #3
    I call your matrix B. Then I presume you solved [itex]\det(B-\lambda\mathbb{I}_{3\times 3})=0[/itex] to find the eigenvalues [itex]\lambda[/itex]. You found [itex]\lambda_1=5 \text{ and } \lambda_2=1[/itex].

    Once you have done that I further assume you tried to find the eigenvector [itex]v_i[/itex] associated to the eigenvector [itex]\lambda_i[/itex] by solving [itex](B-\lambda_i\mathbb{I}_{3\times 3})v_i=0[/itex]. For your eigenvalue one you will find that this equation leads to a two-dimensional solution space for [itex]v_2[/itex], the so called eigenspace. ANY vector in this two-dimensional space will be an eigenvector with eigenvalue one.
    Last edited: Apr 12, 2008
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook