Diagonalization: Am i finding these eigenvectors wrongly?

272 Views Asked by At

$$A=\begin{bmatrix} 1&-2&-8\\ 0&-1&0\\ 0&0&-1 \end{bmatrix}$$ $$P=\begin{bmatrix} 1&-4&1\\ 1&0&0\\ 0&1&0 \end{bmatrix}$$

Confirm that P diagonalizes A. I want to do this by first finding the eigen values and vectors of A then show how these match up with P's columns.

because its an upper triangular $det(A- \lambda I = (1-\lambda)(-1-\lambda)(-1-\lambda)=0$ easily so eigenvalues are 1 and -1 whose multiplicity is 2. Which means -1 must have two free valuables meaning a geometric multiplicity of 2 to be diagonalizable.

So i row reduce $A-I$ and get $$\begin{bmatrix} 0&-2&-8\\ 0&-2&0\\ 0&0&-2 \end{bmatrix}$$ which if if multiplied $R_2$ and $R_3$ by $-\frac{1}{2}$ to get two leading 1s $$\begin{bmatrix} 0&-2&-8\\ 0&1&0\\ 0&0&1 \end{bmatrix}$$ then reducing $R_1$ $$\begin{bmatrix} 0&0&0\\ 0&1&0\\ 0&0&1 \end{bmatrix}$$ I get that $X_1 = t$ so for row 2 $$t+X_2 = 0$$ $$x_2 = -t $$ and like wise $$X_3 = t$$

Which is a vector $t\begin{bmatrix} 1\\ -1\\ -1 \end{bmatrix}$ which doesn't correspond with any column of P even if i change t around for a different value.

Also if i don't multiply $R_{1,2}$ by $-\frac{1}{2}$ and reduce by only multiples of row 2 and row 3 i ge that $$\begin{bmatrix} 0&-0&0\\ 0&-2&0\\ 0&0&-2 \end{bmatrix}$$ in which $X_1 = t \quad X_{2,3}=-\frac{1}{2}t$

$t\begin{bmatrix} 1\\ -\frac{1}{2}\\ -\frac{1}{2} \end{bmatrix}$ which almost make sense weren't for row 1 being a half.

2

There are 2 best solutions below

4
On

The matrix

$$\begin{bmatrix} 0&0&0\\ 0&1&0\\ 0&0&1 \end{bmatrix}$$

Gives $x_1$ as a free variable, and the other two lines give $x_2 = 0$ and $x_3 = 0$. The eigenspace is therefore:

$$\begin{bmatrix} x_1 \\ 0 \\ 0 \end{bmatrix} = x_1\begin{bmatrix}1 \\ 0 \\ 0 \end{bmatrix}$$

2
On

We say that a matrix $P$ diagonalizes a matrix $A$ if $P^{-1} A P$ is diagonal. In our case, however, we have $$ P^{-1} A P = \begin{bmatrix} -1 & 0 & 0\\ 0 & -1 & 0 \\ 0 & -16 & 1 \end{bmatrix} $$ which is not diagonal. This means that the given $P$ does not diagonalize the given $A$.