Scribd
Upload
377 views2 pages

Proof of The Rearrangement Inequality

The rearrangement inequality states that for two sequences of positive real numbers, the sum of their corresponding elements is maximized when the sequences are ordered in the same way (either both ascending or both descending) and minimized when ordered oppositely. The proof involves showing that rearranging one of the sequences cannot increase the sum if the sequences are already ordered the same way. It is proved for two elements and then generalized to any number of elements.

Uploaded by

sahil ahmed
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
377 views2 pages

Proof of The Rearrangement Inequality

The rearrangement inequality states that for two sequences of positive real numbers, the sum of their corresponding elements is maximized when the sequences are ordered in the same way (either both ascending or both descending) and minimized when ordered oppositely. The proof involves showing that rearranging one of the sequences cannot increase the sum if the sequences are already ordered the same way. It is proved for two elements and then generalized to any number of elements.

Uploaded by

sahil ahmed
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

1/19/2018 Proof of the Rearrangement Inequality

Navigation
Home
Proof of the Rearrangement Inequality Contact Graeme
Home
Search Email
Site map Twitter

Math Help > Inequalities > Inequality Methods > Rearrangement Inequality

The Rearrangement Inequality


Let (x) = x1,x2,...,xn and (y) = y1,y2,...,yn be two sequences of positive real numbers. Then the sum

x1y1+x2y2+...+xnyn

is maximized when (x) and (y) are ordered the same way (e.g. both in ascending order or both in
descending order), and the sum is minimized when (x) and (y) are ordered oppositely, with equality iff
(x) is constant or (y) is constant.

Definitions

Two sequences, (x) and (y), are defined as ordered the same way if for any i,j, xi > xj ==> yi ≥ yj and
yi > yj ==> xi ≥ xj,
or equivalently, for all i,j, xi ≤ xj or yi ≥ yj.

Two sequences, (x) and (y), are defined as ordered oppositely if for any i,j, xi > xj ==> yi ≤ yj and yi >
yj ==> xi ≤ xj,
or equivalently, for all i,j, xi ≥ xj or yi ≥ yj.

Proof

When n=2, (x2-x1)(y2-y1) ≥ 0 is equivalent to x1y1+x2y2 ≥ x1y2+x2y1, and (x2-x1)(y2-y1) ≤ 0 is


equivalent to x1y1+x2y2 ≤ x1y2+x2y1.
(x2-x1)(y2-y1) ≥ 0 iff (x) and (y) are ordered the same way, (x2-x1)(y2-y1) ≤ 0 iff (x) and (y) are ordered
oppositely, and (x2-x1)(y2-y1) = 0 if x2=x1 or y2=y1. This completes the proof of the n=2 case.

When n>2, suppose (x) and (y) are ordered the same way, but the sum,

S = x1y1+x2y2+...+xnyn

is not maximized. Then a permutation, (z), of (y) exists such that the sum,

T = x1z1+x2z2+...+xnzn

is maximal, with T larger than S. If (x) and (z) were not ordered the same way, then an i,j would exist
such that xi > xj and zi < zj. In this case, xizi+xjzj < xizj+xjzi (the n=2 case with strict inequality), so
interchanging zi and zj would result in a larger sum, contradicting maximality, so (x) and (z) are
ordered the same way.

Now we have (x) and (y) ordered the same way, giving a sum of S, and (x) and (z) ordered the same
way, giving a larger sum of T>S. But since (y) is a permutation of (z) and (y) and (z) are ordered the
same way, so for all i, yi=zi, and thus T=S.

http://2000clicks.com/mathhelp/IneqRearrangementInequality.aspx 1/2
1/19/2018 Proof of the Rearrangement Inequality

Internet references
The Rearrangement Inequality, by K. Wu, South China Normal University, China and and
Andy Liu, University of Alberta, Canada (mirror)

PlanetMath: Rearrangement Inequality and its proof.

Related pages in this website


The rearrangement inequality is used to prove the Chebyshev Sum Inequality.

The AM-GM Inequality: the Arithmetic Mean of positive numbers is always greater than the
Geometric Mean. This is proved using Jensen's Inequality.

The Cauchy-Schwarz inequality

The Triangle Inequality

Puzzles

The Quadratic Formula

The webmaster and author of this Math Help site is Graeme McRae.

http://2000clicks.com/mathhelp/IneqRearrangementInequality.aspx 2/2

You might also like