3
$\begingroup$

How can I simplify the expression $\sum\limits_{i=1}^n\sum\limits_{j=1}^n x_i\cdot x_j$?

$x$ is a vector of numbers of length $n$, and I am trying to prove that the result of the expression above is positive for any $x$ vector. Is it equal to $\sum\limits_{i=1}^n x_i\cdot \sum\limits_{j=1}^n x_j$? If it is then my problem is solved, because $\left(\sum\limits_{i=1}^n x_i\right)^2$ is non-negative (positive or zero).

$\endgroup$
  • 1
    $\begingroup$ Can you simplify just $\sum_{j=1}^n x_i\cdot x_j$, where $i$ is fixed? $\endgroup$ – Thomas Andrews Dec 7 '14 at 20:57
  • $\begingroup$ I am trying to prove that the result of the expression abouve is positive for any x vector. $\endgroup$ – Serhiy Dec 7 '14 at 21:01
  • 1
    $\begingroup$ Does it equal to $ \sum _{i=1}^n x_{i} \cdot \sum _{j=1}^n x_{j} $ ? If it IS then my problem is solved :) Because $ (\sum _{i=1}^n x_{i} )^2 $ is always positive $\endgroup$ – Serhiy Dec 7 '14 at 21:08
  • 2
    $\begingroup$ @Сергій: Yes. Let $a=\sum_{i=1}^nx_i$. Then $$\sum_{i=1}^n\sum_{j=1}^n(x_ix_j)= \sum_{i=1}^n\left(x_i\sum_{j=1}^nx_j\right)=\sum_{i=1}^nax_i=a^2\;.$$ $\endgroup$ – Brian M. Scott Dec 7 '14 at 21:30
  • 1
    $\begingroup$ @Сергій We know that $\left(\sum\limits_{i=1}^n x_i\right)^2\ge0$ but not necessarily $\left(\sum\limits_{i=1}^n x_i\right)^2>0$. (I would be careful about distinction between the words positive and non-negative.) $\endgroup$ – Martin Sleziak Dec 8 '14 at 6:11
4
$\begingroup$

Yes,

$$\sum_{i=1}^n\sum_{j=1}^nx_i x_j=\left(\sum_{i=1}^ nx_i\right)^2\;.$$

To see this, let $a=\sum_{i=1}^ nx_i$; then

$$\sum_{i=1}^n\sum_{j=1}^nx_i x_j=\sum_{i=1}^n\left(x_i\sum_{j=1}^nx_j\right)=\sum_{i=1}^na x_i=a^2\;.$$

$\endgroup$

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.