Compute determinant of symmetric matrix in general

191 Views Asked by At

Let $$ M = zz^{T},$$ be rank $1$ matrix,where $$ z=\begin{bmatrix} \zeta _{1} \\ \zeta _{2} \\ \vdots \\ \zeta _{n} \end{bmatrix} $$ Compute determinant of $M$.

$M$ is obviously symmetric matrix. Is there any "shortcut" to computing determinant of symmetric matrices in general, or is there no other but to solve this using Laplace expansion?

1

There are 1 best solutions below

0
On BEST ANSWER

One can see it in a different way :

With your notation for a column vector

$$z=\begin{bmatrix} \zeta_{1} \\ \zeta_{2} \\ \vdots \\ \zeta_{n} \end{bmatrix}$$

Let M be a matrix of the form

$$M=[z|v_2|v_3|\cdots | v_n]$$

such that the columns constitute a basis of $\mathbb{R}^n$ (it is always possible to "complete" the basis of a finite dimensional linear space in this way).

Remark : $M$ is therefore invertible.

Then, we can write $$zz^T=MJM^T \ \ \text{where} \ \ J=diag(1,0,0, \cdots 0)$$

(the "1" coefficient selects $zz^T$ and the 0 zero coefficients eliminate any influence of vectors $v_k$).

We can conclude that $rank(zz^T)=rank(J)=1$ because the rank in preserved by pre- or post- multiplication by invertible matrices.

As a consequence $\det(zz^T)=\underbrace{(1 \times 0 \times 0 \cdots \times 0)}_{\det J}\det(M)^2 = 0. :)$