Find eigenvalues for A=[[-4,2,2],[2,-4,2,],[2,2,-4]]
what have you tried so far ?
I've tried finding the determinant of\[A- \lambda I\] but it keeps coming up with complex or decimal values, when it should give -6,-6,0.
i.e. \[\det \left[\begin{matrix}-4- \lambda & 2 & 2 \\ 2 & -4- \lambda & 2 \\ 2 & 2 & -4- \lambda \end{matrix}\right]\]
\[\left|\begin{matrix}-4- \lambda & 2 & 2 \\ 2 & -4- \lambda & 2 \\ 2 & 2 & -4- \lambda \end{matrix}\right| \\ ~\\~ = (-4-\lambda)[(4-\lambda)^2 - 4] - 2[2(-4-\lambda) - 4) + 2[4-2(-4-\lambda)] \\~\\~ \]
Should the second 4 be negative?
\[ = (-4-\lambda)[(4\color{Red}{+}\lambda)^2 - 4] - 2[2(-4-\lambda) - 4) + 2[4-2(-4-\lambda)] \] like this ?
wolfram does give \(\lambda = -6, 0\) : http://www.wolframalpha.com/input/?i=solve++%28-4-%5Clambda%29%28%284%2B%5Clambda%29%5E2+-+4%29-+2%282%28-4-%5Clambda%29+-+4%29+%2B+2%284-2%28-4-%5Clambda%29%29+%3D0
so it seems our characteristic equation is okay, we just need to simplify and solve it...
Why are both the 4 and the lambda positive in that equation (in the brackets where you changed the - to +) when they're negative in the matrix?
because \((-4-\lambda)(-4-\lambda) = (-(4+\lambda))(-(4+\lambda)) = (4+\lambda)(4+\lambda ) = (4+\lambda)^2\)
Okay, that makes sense.
How do we simplify/solve it, because I keep making more complicated and/or incorrect equations.
\[\begin{align} \\ &= (-4-\lambda)[(4\color{Red}{+}\lambda)^2 - 4] - 2[2(-4-\lambda) - 4) + 2[4-2(-4-\lambda)] \\~\\ &= (-4-\lambda)[\lambda^2 + 8\lambda + 12] + 2[2\lambda+8] + 2[12+2\lambda] \end{align}\]
see if that looks okay ^^
Yep, except the third term is missing the minus 4. It should be the same as the last term.
Oh yes i do more mistakes than you haha !
\[\begin{align} \\ &= (-4-\lambda)[(4\color{Red}{+}\lambda)^2 - 4] - 2[2(-4-\lambda) - 4) + 2[4-2(-4-\lambda)] \\~\\ &= (-4-\lambda)[\lambda^2 + 8\lambda + 12] + 2[2\lambda+\color{red}{12}] + 2[12+2\lambda] \end{align}\] looks okay now ?
Haha, I've been through this matrix 5 or 6 times and come up with wrong and different things everytime, you're doing fine :P Looks good.
next may be factor the quadratic in first term
\[\begin{align} \\ &= (-4-\lambda)[(4\color{Red}{+}\lambda)^2 - 4] - 2[2(-4-\lambda) - 4) + 2[4-2(-4-\lambda)] \\~\\ &= (-4-\lambda)[\lambda^2 + 8\lambda + 12] + 2[2\lambda+\color{red}{12}] + 2[12+2\lambda] \\~\\ &= (-4-\lambda)[(\lambda +6)(\lambda+2)] + 4[\lambda+\color{red}{6}] + 4[6+\lambda] \\~\\ \end{align}\]
I've expanded it out to get\[-48-32\lambda -4\lambda ^{2}-12\lambda +8\lambda ^{2}-\lambda ^{3}+48-8\lambda\] Does that look right?
It simplifies to \[\lambda ^{3}+12\lambda ^{2} +36\lambda\] \[=\lambda(\lambda ^{2}+12 \lambda +36)\]
\[=\lambda (\lambda +6)(\lambda +6)\]
So lambda = -6, -6, 0 Which is the answer!
looks perfect !
Thank you so much, I was so stuck!
np :) also there is an interesting property for eigen values : they add up to the sum of elements in diagonal
\[ \left[\begin{matrix} \color{Red}{-4} & 2 & 2 \\ 2 & \color{Red}{-4} & 2 \\ 2 & 2 & \color{Red}{-4} \end{matrix}\right] \]
\(\large \color{Red}{-4-4-4} = -6-6+0 = -12\)
Thanks, that'll help with checking my answers!
this is in general true for all matrices
if you notice this is really a special matrix : symmetrical matrix with constant diagonal
also this is singular - thats the reason we have got 0 as one of the eigen values
Join our real-time social learning platform and learn together with your friends!