1
$\begingroup$

Suppose there are 2n vectors $\{m_1,m_2,...,m_n\}$ and $\{\mu_1,\mu_2,...,\mu_n\}$. All vectors are in k-dimensional Euclid space $R^k$. The vectors satisfy: $$m_i \mu_j \leq 0, \quad \forall i\neq j $$ and $$ m_i \mu_i > 0, \quad \forall i.$$

What I want to prove: $n \leq 2k$.

Intuitively I think the bound $n \leq 2k$ is correct. It can be achieved by setting $m_i$ and $\mu_i$ to be the canonical base vectors and their negatives.

As Saul points out, the upper bound actually does not exist in this simplified problem. I post my Orginal Problem in a new post.

$\endgroup$
4
  • 1
    $\begingroup$ $n=k=2$, $m_1=\mu_1=(1,0)$, $m_2=\mu_2=(-1,0)$ is a counterexample. $\endgroup$ Commented Jan 29, 2023 at 22:18
  • 1
    $\begingroup$ Using the canonical basis vectors and their negatives will give you $n = 2k$. Should the inequality be $n \le 2k$? $\endgroup$ Commented Jan 30, 2023 at 0:35
  • 1
    $\begingroup$ Of course, @GerryMyerson you are correct. It was typo, I want to prove $n\leq 2k$, as Robert suggested, not $2n \leq k$. Sorry for the confusion. $\endgroup$
    – TanG
    Commented Jan 30, 2023 at 15:08
  • 1
    $\begingroup$ Yes, you are correct. My apology :) @GerryMyerson $\endgroup$
    – TanG
    Commented Jan 30, 2023 at 15:12

1 Answer 1

2
$\begingroup$

You can pack as many vectors $m_i$ and $\mu_i$ as you want in $\mathbb{S}^2=\{x^2+y^2+z^2=1\}\subseteq\mathbb{R}^3$. I leave below an example with $7$ points that generalizes to any finite number of points.

Consider points $m_1,\dots,m_7$ forming a regular heptagon in $\mathbb{S}^2\cap\{z=-0.01\}$. Below is a figure showing how a small neighborhood of the north pole $N=(0,0,1)$ in $\mathbb{S}^2$ would look like, with the lines being the sets $\{v\in\mathbb{S}^2;\langle m_i,v\rangle=0\}$, for $i=1,\dots,7$. Note that $\langle m_i,N\rangle<0\;\forall i$.

enter image description here

Now choose $\mu_i$ in the positions of the seven points from the figure, so that $\langle \mu_i,m_i\rangle>0$ for all $i$ but for $j\neq i$ we have $\langle \mu_i,m_j\rangle<0$.

$\endgroup$
4
  • 2
    $\begingroup$ if a question is correctly stated and answered, it is usually considered "best-practice" to accept the answer. then you can ask a more technical version of your question in a new post. $\endgroup$ Commented Feb 11, 2023 at 20:38
  • $\begingroup$ Thank you Saul! A very nice example. $\endgroup$
    – TanG
    Commented Feb 11, 2023 at 22:22
  • $\begingroup$ @TanG You are welcome :) I was in the middle of editing my answer. Are you planning on posting a new question in the end? $\endgroup$
    – Saúl RM
    Commented Feb 11, 2023 at 22:25
  • $\begingroup$ @SaúlRM Thanks. Yes, I just post a new question. I've added the link in my question. $\endgroup$
    – TanG
    Commented Feb 11, 2023 at 23:45

You must log in to answer this question.

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