4
$\begingroup$

Suppose that $G=(G_{ij})$ is a positive-semidefinite symmetric matrix with the diagonal entries all equal $1$ and all off-diagonal entries $\le0$. Does it then necessarily follow that $$\sum_{i,j}(G^5)_{ij}\le\sum_{i,j}(G^3)_{ij}\,?$$

This is true if it is additionally assumed that all off-diagonal entries of $G$ are the same.

$\endgroup$

1 Answer 1

9
$\begingroup$

If I have not committed any mistake, please, find below a counter-example.

Counter-example. Let $G\in \mathbb{S}^3_{+}$ be defined by $$ G = \begin{pmatrix} 1 & -\frac{2}{5} & 0 \\ -\frac{2}{5} & 1 & -\frac{2}{5}\\ 0 & -\frac{2}{5} & 1 \end{pmatrix}. $$

$\endgroup$
4
  • 2
    $\begingroup$ $\frac{2}{5}$ replaced with $\frac{1}{2}$ works too. $\endgroup$ Jun 3, 2023 at 11:27
  • $\begingroup$ Mark, thank you for the edit. @PeterMueller, indeed. And it works for 1/3 as well, but it fails (i.e., the inequality holds true), e.g., for $1/5$. Not clear at this point for me what would be a 'good' sufficient condition for this inequality to hold. $\endgroup$ Jun 3, 2023 at 12:50
  • $\begingroup$ Thank you for your answer. $\endgroup$ Jun 4, 2023 at 3:09
  • $\begingroup$ @IosifPinelis, thank you. $\endgroup$ Jun 4, 2023 at 12:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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