Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

The following statement seems clearly true, but I'm having a hard time proving it:

Fix $\alpha\in[0,1]$. Let $\mu$ be Lebesgue measure. For $r\ge 0$, let $B(c,r)\equiv[c-r,c+r]$. Fix $r_1,\ldots,r_n\in[0,\infty)$, and $c_1,\ldots,c_n\in\mathbb R$. Then, $\mu\left(\bigcap_{i=1}^{n}B(\alpha c_{i},r_{i})\right)\ge\mu\left(\bigcap_{i=1}^{n}B(c_{i},r_{i})\right)$

In words, the statement just says that contracting the 1-dimensional balls' centers towards $0$, by a constant proportion, while leaving the radii unchanged, increases the amount of overlap. This seems straightforward enough: if the centers were contracted all the way to the origin, then the intersection would just be the smallest ball. There can't be more overlap than that.

Can anyone suggest a simple proof, or point me to a relevant theorem?

share|improve this question

2 Answers 2

up vote 1 down vote accepted

Let $B(c,r)=\bigcap_{i=1}^n B(c_i,r_i)$. We have $|c_i-c|\le r_i-r$ for all $i$. Hence, $|\alpha c_i-\alpha c|\le r_i-r$, which implies $B(\alpha c,r)\subseteq \bigcap_{i=1}^n B(\alpha c_i,r_i)$.

Explanation: a ball $B$ is contained in ball $B'$ if and only if the distance between centers is at most radius(B')-radius(B).

share|improve this answer
    
Here it helps that in 1 dimension the intersection of balls is also a ball. I wonder what a proof would be like in higher dimensions. As far as I know, for general 1-Lipschitz maps $f\colon \mathbb R^n\to\mathbb R^n$ the inequality $\mu\left(\bigcap_{i=1}^{n}B(f(c_{i}),r_{i})\right)\ge\mu\left(\bigcap_{i=1}^{n}‌​B(c_{i},r_{i})\right)$ remains conjectural. –  user31373 Jun 21 '12 at 21:12
    
Thanks so much! –  Jeff Jun 21 '12 at 23:05

Consider the case of two intervals. Once we fix the radii, the volume of the intersection is indeed a decreasing function of the distance between the centres, so the result is straightforward in this case. But an intersection of intervals is either another interval or is empty, so really this is all there is to it. If you like, you have the result by induction.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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