0
$\begingroup$

For any square matrix A,

  1. $A^tA$ is definite positive, since $x^tA^tAx=y^ty>0, \forall x\in\mathbb{R}^n≠0$ with $y=Ax$.
  2. Definite positive matrices are invertible. (If a matrix B was singular, there would exist $x\in\mathbb{R}^n ≠ 0$ such that $Bx = 0$. But then $Bx = 0 \Rightarrow x^tBx = 0$, which is a contradiction. So B must be invertible.) So $det(A^t A) ≠ 0$.
  3. $det(A^t A)=det(A^t) det(A)$ since $A^t$ and $A$ are square matrices of the same size. So if $det(A^t A)≠0$ then $det(A)≠0.$
  4. Therefore, A is invertible.

I just can't see it...

$\endgroup$
2
  • 2
    $\begingroup$ Just try it on $A=\begin{bmatrix}0 & 0\\0 & 0\\\end{bmatrix}$ and you will see where you went wrong. $\endgroup$
    – TonyK
    Commented Dec 10, 2022 at 15:04
  • $\begingroup$ as pointed out by @Anne Bauval already...just because $x\neq 0$ does not mean you will always have $Ax\neq 0$. $\endgroup$
    – Rescy_
    Commented Dec 10, 2022 at 15:04

1 Answer 1

2
$\begingroup$

It is your point 1 which is wrong. $A^t A$ is generally only positive semi-definite, i.e. it may happen that $x^tA^tAx=0,$ i.e. $Ax=0,$ for some non-zero vectors $x.$

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .