I'm studying for my exam of linear algebra.. I want to prove the following corollary:

If $A$ is a symmetric positive definite matrix then each entry $a_{ii}> 0$, ie all the elements of the diagonal of the matrix are positive.

My teacher gave a suggestion to consider the unit vector "$e_i$", but I see that is using it.

$a_{ii} >0$ for each $i = 1, 2, \ldots, n$. For any $i$, define $x = (x_j)$ by $x_i =1$ and by $x_j =0$, if $j\neq i$, since $x \neq 0$, then:

$0< x^TAx = a_{ii}$

But my teacher says my proof is ambiguous. How I can use the unit vector $e_1$ for the demonstration?

  • 1
    $\begingroup$ Please use LaTeX formatting (I have a pending edit that formatted it for you). I am not sure what you mean by xj 00. $\endgroup$ – Emily Aug 8 '12 at 4:14
  • 1
    $\begingroup$ Your proof seems OK, generally speaking; I suspect that your teacher objected to the presentation. It doesn't make sense to start out with $a_{ii}\gt0$; this is what you want to prove, not something you're starting from. $\endgroup$ – joriki Aug 8 '12 at 4:22
  • $\begingroup$ @EdGorcenski Sorry is xj =0. $\endgroup$ – franvergara66 Aug 8 '12 at 4:25

Let $e_1 = \begin{pmatrix} 1 \\ 0 \\ \vdots \\ 0 \end{pmatrix}$, and so on, where $e_i$ is a vector of all zeros, except for a $1$ in the $i^{\mathrm{th}}$ place. Since $A$ is positive definite, then $x^T A x > 0$ for any non-zero vector $x \in \Bbb R^n$. Then, $e_1^T A e_1 > 0$, and likewise for $e_2, e_3$ and so on.

If the $i^{\mathrm{th}}$ diagonal entry of $A$ was not positive, $a_{ii} < 0$, then $e_i^T A e_i = 0\cdot a_{11}\cdot 0 + 1\cdot a_{12}\cdot 0 + \cdots + 1\cdot a_{ii}\cdot 1 + \cdots + 0\cdot a_{nn} \cdot 0$, since $e_i$ has zeros everywhere but in the $i^{\rm th}$ spot.

Thus, what would happen if $a_{ii}$ was positive?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.