• Nem Talált Eredményt

A STABILITY VERSION OF HÖLDER’S INEQUALITY FOR 0 < p < 1

N/A
N/A
Protected

Academic year: 2022

Ossza meg "A STABILITY VERSION OF HÖLDER’S INEQUALITY FOR 0 < p < 1"

Copied!
7
0
0

Teljes szövegt

(1)

Hölder’s Inequality J.M. Aldaz vol. 9, iss. 2, art. 60, 2008

Title Page

Contents

JJ II

J I

Page1of 7 Go Back Full Screen

Close

A STABILITY VERSION OF HÖLDER’S INEQUALITY FOR 0 < p < 1

J. M. ALDAZ

Departamento de Matemáticas y Computación Universidad de La Rioja

26004 Logroño, La Rioja, Spain.

EMail:jesus.munarrizaldaz@unirioja.es

Received: 21 October, 2007

Accepted: 23 March, 2008

Communicated by: L. Losonczi 2000 AMS Sub. Class.: 26D15.

Key words: Hölder’s inequality, Reverse triangle inequality.

Abstract: We use a refinement of Hölder’s inequality for1< p < to obtain the cor- responding refinement whenr (0,1). This in turn allows us to sharpen the reverse triangle inequality on the nonnegative functions inLr, forr(0,1).

Acknowledgements: The author was partially supported by Grant MTM2006-13000-C03-03 of the D.G.I. of Spain.

(2)

Hölder’s Inequality J.M. Aldaz vol. 9, iss. 2, art. 60, 2008

Title Page Contents

JJ II

J I

Page2of 7 Go Back Full Screen

Close

BykFkt := R

|F|t1/t

we do not mean to imply that this quantity is finite, nor do we assume thatt >0; in fact, in this note negative exponents are unavoidable.

It is well known that Hölder’s inequality can be extended to the range0< r <1, by an argument that essentially amounts to a clever rewriting of the case1< p <∞, cf. [2, pg. 191]. We denote the conjugate exponent ofrbys :=r/(r−1), and the conjugate exponent ofp byq := p/(p−1)(of course, to go from the range (0,1) to (1,∞) and viceversa, one sets r = 1/p). Hölder’s inequality for 0 < r < 1 tells us that if h and k are nonnegative functions in Lr and Ls respectively, then R hk ≥ R

hr1/r R ks1/s

. This entails that given functions h, w ≥ 0 inLr, the reverse triangle inequality kh+wkr ≥ khkr +kwkr holds. Nonnegativity is of course crucial.

Here we extend to the range (0,1) the following stability version of Hölder’s inequality, which appears in [1]:

Let 1 < p < ∞ and letq = p/(p−1)be its conjugate exponent. If f ∈ Lp, g ∈Lqare nonnegative functions withkfkp,kgkq >0, and1< p≤2, then

(1) kfkpkgkq 1− 1 p

fp/2

kfp/2k2 − gq/2 kgq/2k2

2

2

!

+

≤ kf gk1 ≤ kfkpkgkq 1− 1 q

fp/2

kfp/2k2 − gq/2 kgq/2k2

2

2

! ,

while if2≤p < ∞, the terms1/pand1/qexchange their positions in the preceding inequalities.

Inequality (1) essentially states that kf gk1 ≈ kfkpkgkq if and only if the angle between theL2vectorsfp/2andgq/2is small (in this sense it is a stability result). To see that on the cone of nonnegative functions (1) extends the parallelogram identity,

(3)

Hölder’s Inequality J.M. Aldaz vol. 9, iss. 2, art. 60, 2008

Title Page Contents

JJ II

J I

Page3of 7 Go Back Full Screen

Close

rearrange the latter, for nonzerox andy in a real Hilbert space, as follows (cf. [1, formula (2.0.2)]):

(2) (x, y) = kxkkyk 1−1 2

x

kxk− y kyk

2! .

Writing (2) as a two sided inequality, adequately replacing some of the Hilbert space norms bypandqnorms, and the terms1/2by1/pand1/q, we see that (1) indeed generalizes (2). Note also thatkfp/2k2 =kfkp/2p . Save in the case wherep=q = 2, the nonnegative functions f ∈ Lp andg ∈ Lq will in principle belong to different spaces, so to compare themL2is retained in (1) as the common measuring ground; to go fromLp andLq intoL2 we use the Mazur map, which for nonnegative functions of norm 1 inLp is simplyf 7→fp/2(cf. [1] for more details).

Next we extend inequality (1) to the range 0 < r < 1, keeping the role of L2. Unlike the case of Hölder’s inequality for1< p <∞, here we assume thathk ∈L1. In exchange, we do not need to suppose a priori thath ∈Lr; this will be part of the conclusion.

Theorem 1. Let0 < r < 1, and let s = s/(s−1)be its conjugate exponent. If k ∈Ls,hk ∈L1,khkr,kkks >0, and1/2≤r <1, then

(3a) khkk1 1−r

h1/2k1/2

kh1/2k1/2k2 − ks/2 kks/2k2

2

2

!1r

+

(3b) ≤ khkrkkks ≤ khkk1 1−(1−r)

h1/2k1/2

kh1/2k1/2k2 − ks/2 kks/2k2

2

2

!1r , while if0< r≤1/2, the termsrand1−rexchange their positions in the preceding inequalities.

(4)

Hölder’s Inequality J.M. Aldaz vol. 9, iss. 2, art. 60, 2008

Title Page Contents

JJ II

J I

Page4of 7 Go Back Full Screen

Close

Proof. Suppose1/2≤ r < 1. Setp= 1/rand useq ands to denote the conjugate exponents ofpandrrespectively. Since1< p≤2, we can apply (1) to the functions f :=hrkrandg =k−r, which belong toLpandLq respectively:R

fp =R

hk < ∞ andR

gq =R

ks<∞. Now the inequalities (3) immediately follow. If0< r≤1/2, then2≤p <∞, so just interchange the terms1/pand1/qin (1).

Note that from (3b), together with the hypothesiskhkrkkks>0, we get

(4) 0<1−(1−r)

h1/2k1/2

kh1/2k1/2k2 − ks/2 kks/2k2

2

2

for allr ∈ [1/2,1)(forr ∈ (1/2,1)this already follows from

x

kxk2kyky

2

2 2

≤ 2, which is immediate from (2) whenx, y ≥ 0). The analogous result, withr instead of1−r, holds when0< r≤1/2. Thus, (3b) can be rewritten as

(5) khkrkkks 1−(1−r)

h1/2k1/2

kh1/2k1/2k2 − ks/2 kks/2k2

2

2

!1r

≤ khkk1

when 1/2 ≤ r < 1, while if 0 < r ≤ 1/2, the same formula holds but with r replacing1−r.

Now we are ready to obtain a sharpening of the reverse triangle inequality for nonnegative functions.

(5)

Hölder’s Inequality J.M. Aldaz vol. 9, iss. 2, art. 60, 2008

Title Page Contents

JJ II

J I

Page5of 7 Go Back Full Screen

Close

Theorem 2. Let0< r <1. Given nonnegative functionsh, w∈Lrwithkhkr,kwkr >

0, setk := (h+w)r−1/k(h+w)r−1ks. Then, if1/2≤r <1, we have

(6) kh+wkr ≥ khkr 1−(1−r)

h1/2k1/2

kh1/2k1/2k2 −ks/2

2

2

!1r

+kwkr 1−(1−r)

w1/2k1/2

kw1/2k1/2k2 −ks/2

2

2

!1r

,

while if0< r≤1/2, the same inequality holds but with1−rreplaced byr.

Proof. Suppose1/2≤r <1, and note thatkis a unit vector inLs. Hence, so isks/2 inL2. By the nonnegativity ofhandwwe have

(7) kh+wkr=

Z (h+w)r−1

k(h+w)r−1ks(h+w) = Z

hk+ Z

wk.

Since the left hand side of the preceding equality is finite, so are both integrals on the right hand side, and now the result follows by applying (4). If0< r ≤ 1/2,we argue in the same way, but withrreplacing1−rin (4).

Let us writeθ(x, y) :=

x

kxkkyky

. To conclude, we make some comments on the size ofθ(h1/2k1/2, ks/2), which also apply toθ(w1/2k1/2, ks/2). On a real Hilbert space,θ(x, y)is comparable to the angle between the vectorsxandy. In particular, θ(h1/2k1/2, ks/2)is zero if and only if there exists at >0such thath=tw, in which casekh+wkr =khkr+kwkr. Under any other circumstance, the inequality given by (6) is strictly better that the standard reverse triangle inequality.

(6)

Hölder’s Inequality J.M. Aldaz vol. 9, iss. 2, art. 60, 2008

Title Page Contents

JJ II

J I

Page6of 7 Go Back Full Screen

Close

On the other hand, if we ask how small 1−(1−r)

h1/2k1/2

kh1/2k1/2k2 −ks/2

2

2

!1r

can be forr ∈ [1/2,1),the obvious bound θ(h1/2k1/2, ks/2) ≤ √

2 is informative whenris close to 1, but useless ifr= 1/2. The analogous remark holds for

1−r

h1/2k1/2

kh1/2k1/2k2 −ks/2

2

2

!1r

when0 < r ≤ 1/2. However, nontrivial bounds also hold near1/2, since for every r∈(0,1),kh+wkr≤21/r−1(khkr+kwkr)(see for instance Exercise 13.25 a), [2, pg. 199]). Thus, θ(h1/2k1/2, ks/2)and θ(w1/2k1/2, ks/2) cannot be simultaneously large. More precisely, if1/2≤r <1, then either

θ2(h1/2k1/2, ks/2)≤ 1−2r−1 1−r or

θ2(w1/2k1/2, ks/2)≤ 1−2r−1 1−r , while if0< r≤1/2, then either

θ2(h1/2k1/2, ks/2)≤ 1−2r−1 r or

θ2(w1/2k1/2, ks/2)≤ 1−2r−1

r .

(7)

Hölder’s Inequality J.M. Aldaz vol. 9, iss. 2, art. 60, 2008

Title Page Contents

JJ II

J I

Page7of 7 Go Back Full Screen

Close

References

[1] J.M. ALDAZ, A stability version of Hölder’s inequality, J. Math. Anal.

Applics., to appear. doi:10.1016/j.jmaa.2008.01.104. [ONLINE: http://

arxiv.org/abs/0710.2307].

[2] E. HEWITTANDK. STROMBERG, Real and Abstract Analysis. A Modern Treatment of the Theory of Functions of a Real Variable. Second printing corrected, Springer-Verlag, New York-Berlin, 1969.

Hivatkozások

KAPCSOLÓDÓ DOKUMENTUMOK

We give a reverse inequality to the most standard rearrangement inequality for sequences and we emphasize the usefulness of matrix methods to study classical inequalities.. Key

Thus, in a sense, the arithmetic-geometric mean inequality, Hölder’s inequality, the Cauchy-Schwarz inequality, and Lyapunov’s inequality are all equivalent [1, p. In the fol-

Key words and phrases: Hilbert’s inequality, Hölder’s inequality, Jensen’s inequality, Power mean inequality.. 2000 Mathematics

In recent years, [2] – [5] consid- ered the strengthened version, generalizations and improvements of inequality (1.1) and Pach- patte [6] built some inequalities similar to

Key words: Triangle inequality, Reverse inequality, Diaz-Metkalf inequality, Inner product

Key words: Schatten class, Schatten norm, Norm inequality, Minkowski inequality, Triangle inequality, Powers of operators, Schatten-Minkowski constant.. This work was supported by

THOMPSON, The case of equality in the matrix-valued triangle inequality, Pacific J.. ZENG, Young’s inequality in compact

Peˇcari´c [2] established a new class of related integral inequalities from which the results of Pachpatte [12] – [14] are obtained by specializing the parameters and the functions