- Research
- Open access
- Published:
Points of nonsquareness of Lorentz spaces
Journal of Inequalities and Applications volume 2014, Article number: 467 (2014)
Abstract
Criteria for nonsquare points of the Lorentz spaces of maximal functions are presented under an arbitrary (also degenerated) nonnegative weight function w. The criteria for nonsquareness of Lorentz spaces and of their subspaces of all order continuous elements, proved directly in (Kolwicz and Panfil in Indag. Math. 24:254-263, 2013), are deduced.
MSC: 46E30, 46B20, 46B42.
1 Introduction
The geometry of Banach spaces has been intensively developed during the last decades. Nonsquareness and uniform nonsquareness are important properties in this area. Uniform nonsquareness implies both superreflexivity and the fixed point property (see [1, 2] and [3]). Therefore it is natural to investigate nonsquareness properties in various classes of Banach spaces (see [4–10]). They are also connected with the notion of James constant (see [11–13]) which describes the measure of nonsquareness. The class of uniformly nonsquare Banach spaces is strictly smaller than the class of B-convex Banach spaces. Recall that B-convexity plays an important role in the probability (see [14]).
On the other hand, it is natural to ask whether a separated point x in a Banach function space E has some local property P whenever the whole space E does not possess this property. This leads to the local geometry which has been deeply studied recently (see [15–19]). The monotonicity properties of separated points have applications in best dominated approximation problems in Banach lattices (see [15]). The extreme points, and SU points play a similar role in the theory of Banach spaces.
The purpose of this paper is to characterize nonsquare points of the Lorentz space . We also give a criterion for a point to be nonsquare in the subspace of order continuous elements of . Since degenerated weight functions w are admitted, such investigations concern the most possible wide class of these spaces. Moreover, the local approach presented in this paper required new techniques and methods (in comparison with the global approach in [20]), which may be of independent interest.
2 Preliminaries
Let ℝ be the set of real numbers and or be the unit sphere of a real Banach space X. Denote by the set of all m-equivalence classes of real-valued measurable functions defined on with m being the Lebesgue measure on ℝ and or .
A Banach lattice is called a Banach function space (or a Köthe space) if it is a sublattice of satisfying the following conditions:
-
(1)
If , and a.e., then and .
-
(2)
There exists a strictly positive on , .
The symbol stands for the positive cone of E, that is, . We say that E has the Fatou property if for any sequence such that for all , , a.e. with , we have and .
We say that a Banach function space is rearrangement invariant (r.i. for short) if whenever and with , then and (see [21]). Recall that stands for the distribution function of , that is, for every . Then the nonincreasing rearrangement of f is defined by
for . Given , we denote the maximal function of by
It is well known that and for any (see [22, 23] for other properties of and ).
A Banach function space E is said to be strictly monotone () if for each with we have . A point is a point of lower monotonicity (upper monotonicity) if for any such that and (respectively, and ), we have (respectively, ). We will write shortly that f is an LM point and UM point, respectively.
Clearly, the following assertions are equivalent:
-
(i)
E is strictly monotone (shortly, );
-
(ii)
each point of is a point of upper monotonicity;
-
(iii)
each point of is a point of lower monotonicity.
Definition 2.1 Let and or . Let be a nonnegative weight function. The Lorentz space is a subspace of functions satisfying the following formula:
Throughout the paper, we assume that w satisfies the following conditions:
for all if and for all in the opposite case. These two conditions assure that and is a rearrangement invariant Banach function space with the Fatou property (see [24, 25]).
These spaces were introduced by Calderón in [26] and are naturally related to classical Lorentz spaces defined by Lorentz in [27]. Obviously, for any and these spaces coincide if and only if the Hardy operator is bounded on . This condition is equivalent to the so-called condition related to the weight w (see [25, 28–30]). It is also worth mentioning that spaces appear naturally in the interpolation theory as a result of the Lions-Peetre K-method. These spaces have been recently intensively investigated from both the isomorphic as well as the isometric point of view (see [24, 25, 31]).
Definition 2.2 A point is a point of nonsquareness (we write shortly f is an NSQ point) provided that
for all . A Banach space is nonsquare ( for short) if each point of is an NSQ point.
Notation 2.1 For simplicity, we will sometimes use the following notations:
-
(a)
.
-
(b)
By we denote the support of .
-
(c)
if .
-
(d)
if .
-
(e)
For measurable subsets A, B of ℝ, by we mean .
Let us recall some useful properties of a nonincreasing rearrangement operator.
Lemma 2.1 ([23], Property 7∘, p.64)
Let . If , then there is a set with and
Remark 2.1 Let with or . The above lemma holds (without the assumption )
-
(a)
for every , where ;
-
(b)
for every in the case of .
Lemma 2.2 ([23], Property 8∘, p.64)
The equality
holds for .
Remark 2.2 Lemma 2.2 implies
i.e., the subadditivity property of the maximal function.
Remark 2.3 Let . The inequality
for with implies . Indeed,
The following result is a generalization of Lemma 1 from [20].
Lemma 2.3 Let . If , then
for every .
Proof Set
with the convention .
Since , so . Assume, without loss of generality, that . Clearly,
Notice that for every if and for every if , by Lemma 1 in [20], we have
Moreover, if , then for ,
since for every .
If then, by (3) and (4), we get
for .
If then, by (2) and (4), we have
for . □
Remark 2.4 Let and . Then is constant on if and only if .
Proof Clearly, and for all . Therefore is equivalent to for some . □
Theorem 2.1 Let E be a symmetric Banach function space, and . If x is an NSQ point, then .
Proof Assume . By Remark 2.4, for every . Thus for every and for every . Moreover, every function of the type with is equimeasurable with x since for every and for every . Therefore .
Denote
Case I. Suppose and denote
If , then take and such that , and . Define
Thus
whence , i.e., x is not an NSQ point. The case goes analogously.
Case II. Let and . Define
Note that either for all or for all , since in the opposite case there are that
Thus, taking , we get
whence
a contradiction.
Without loss of generality, we may assume for all . For , let
and note , for all and . Let , satisfy , , and . Denote
Notice , and . We claim that
for all . If there exists such that , then take satisfying . Thus
Therefore , a contradiction. The case is analogous, which proves claim (5).
Let . Then, for all , we have
where , since and . Analogously, for , .
Obviously, by assumption that , and . Thus
whence for every . Therefore , whence x is not an NSQ point. □
In the sequel we will use the following notations:
with the convention , .
Theorem 2.2 Let . If x is an NSQ point, then
-
(i)
is not constant on if ,
-
(ii)
is not constant on if and ,
where γ is defined in (6).
Proof The case follows from Remark 2.4 and Theorem 2.1.
Consider if or if . For the contrary, assume that for some . Let . Then, in the case of , Theorem 2.1 implies . Thus . By Lemma 2.1, there are disjoint sets and , both of measure γ, such that and . Moreover, and for . Taking , we get and . Since , then , i.e., x is not an NSQ point. □
Theorem 2.3 If is an NSQ point, then , where β is defined in (6).
Proof Suppose . This means . Take , , where , , and . Then and since . □
Theorem 2.4 Let , β and γ are as in (6), and let the weight function be such that . The function x is an NSQ point if and only if and is not constant on .
Proof Necessity. It follows from Theorems 2.2 and 2.3.
Sufficiency. Let . If then, by Lemma 2.3, for all . Since , so , whence (see Remark 2.3 and the definition of β), i.e., x is not an NSQ point.
Now assume . Denote
We have
Obviously,
We will consider the following pairwise independent cases.
-
Case I. .
-
Case II. .
-
Case II.A. There is such that , or there is with .
-
Case II.B. For every , .
-
Now let us discuss all the cases.
Proof of Case I. Since and , then satisfies the conditions (i) and (ii) of Theorem 3.1 in [15], whence is an LM point. Moreover, by (8), at least one of the following inequalities holds:
Since , one of the following inequalities holds:
Proof of Case II.A. Assume that there exists such that . Since , so . By the right continuity of nonincreasing rearrangement function, there exists such that for all . Therefore,
for . It is clear that (see Remark 2.2), whence
for every . Since , so (see Remark 2.3).
Notice, if there is such that , then analogous reasoning gives .
Proof of Case II.B. Assume
for every . Consequently,
Denote
(a) Suppose . Then
By (12), there is satisfying
for every . An analogous inequality holds for and . Take the sets , , of measure such that
(see Lemma 2.1). Clearly, by the definition of and equimeasurability of a function and its nonincreasing rearrangement, we get
Let
(see notation (7)). By , (10), (11) and (13),
Analogously, the equality gives
and yields
We claim that . Indeed, if , then by Lemma 2.2 and (10), we get
a contradiction. Analogous reasoning goes for , which proves the claim.
Moreover, the above arguments and (9) imply , since
and
Thus , whence (15) implies . Summarizing, we get
For there exists a set of measure t such that
where
We have and, by (16), . By the above argumentation, together with Lemma 2.2 and , at least one of the following inequalities holds:
for . Thus, for every ,
By (12), we may find a sequence , , such that inequality (13) is satisfied for each . Similarly as above, we conclude inequality (17) with instead of . Consequently, (17) holds for all . This means (see Remark 2.3).
(b) Assume and take the sets , , of measure such that
(see Lemma 2.1). Clearly, by the definition of and equimeasurability of a function and its nonincreasing rearrangement, we get
Let , and for be defined as in (14). By , (10) and (11) we conclude (15). Moreover, similarly as above, we get (17) for instead of .
Moreover, by the definition of , for every ,
Finally, by (17) and the above inequality, we get
for every . Therefore, (see Remark 2.3).
(c) Suppose , i.e.,
for some . Notice . Then, for every , we have
Additionally, for all ,
Since and satisfies the conditions (i) and (ii) of Theorem 3.2 in [15], so is a UM point. Thus . Therefore,
which finishes the proof. □
Theorem 2.5 An element is an NSQ point if and only if and, if , is not constant on , where β and γ are defined in (6).
Proof Necessity. It follows from Theorems 2.2 and 2.3.
Sufficiency. Let . If x and y have disjoint supports, then, by Lemma 1 in [20], for all . Since , so , whence (see Remark 2.3).
Assume that x and y have not disjoint supports, i.e.,
where
. Assume . By Theorem 3.1 in [15], is an LM point. By (18), at least one of the inequalities holds:
Obviously, , whence at least one of the inequalities holds: or .
ℬ. Suppose
Denote the sets , , , , of measure γ satisfying
(see Lemma 2.1 and Remark 2.1). Notice
Moreover,
Consider the following cases.
-
Case I. There is such that , or there is such that .
-
Case II. for all .
-
Case II.1. or .
-
Case II.1.A. or .
-
Case II.1.B. and .
-
-
Case II.2. and .
-
Now let us discuss all the cases.
Proof of Case I. Assume that there is such that . Since so . By the right continuity of nonincreasing rearrangement, there is such that for all . Therefore,
for . It is clear that (see Remark 2.2), whence
for every . By the definition of γ, , so (see Remark 2.3).
Notice, if there is that , then analogous reasoning gives .
Proof of Case II. Suppose
for every . Condition (23) implies
since otherwise if, for example, , then, by ,
a contradiction with (23).
Notice
by (20) and (23).
Case II.1. Assume . Then implies
By (24), we have
Case II.1.A. Assume . Then by (26), and since . Thus
Moreover, by (20) and , we get .
We claim that
Assume for the contrary that . By (27) and (28),
Moreover, by (24) and , we get . Furthermore, by (21) and (23), we get
a contradiction with (30). This proves claim (29). Therefore, (27), (28), (29) imply , which finishes the proof (see Remark 2.3 and the definition of γ).
It is clear that analogous reasoning holds for the case of .
Case II.1.B. Assume and . Then
We claim that at least one of inequalities (32) or (33) holds,
If or , then, by (26) and (31), we get (32) or (33), respectively.
If and , then
Assume for the contrary that (32) and (33) do not hold, i.e.,
The equality and (24) imply
and analogously we get
Therefore, by (35), we have
and
Thus by (34), and , whence . Since , we get a contradiction with (20). This proves that (32) or (33) holds.
Finally, by (27) and (32) or (33), we get , which finishes the proof (see Remark 2.3 and the definition of γ).
Considering the case of , we may follow analogously but with element .
Case II.2. Suppose and . Then, by (24),
We claim that and . If , then , whence by the definition of set , a contradiction with (25). The case of goes analogously and proves the claim.
By (22), and . Since is not constant on , then
Conditions (36) imply . Consequently,
Thus
Finally, by (36), one of the following holds:
or
which finishes the proof (see Remark 2.3 and the definition of γ). □
Below we present some modification of Lemma 2.1 from [15].
Lemma 2.4 Let satisfy , for , and for every . Then there is a set B of positive measure such that for .
Proof Denote . The case of is done in Lemma 2.1 in [15]. Assume . Thus . Define
where , .
Then and x as well as and y are equimeasurable. Since and satisfy the assumptions of Lemma 2.1 in [15], there is a set B with such that for . □
Theorem 2.6 Let , β and γ are as in (6), and let the weight function be such that . Then x is an NSQ point if and only if and is not constant on .
Proof Necessity. It follows from Theorems 2.2 and 2.3.
Sufficiency. Let . If , then by Lemma 2.3, for all . Since , so , whence (see Remark 2.3 and the definition of β).
Denote
and assume
Obviously,
. Assume . We follow as in the proof of Theorem 2.5, Case .
ℬ. Suppose
Consider the following cases (for the definitions of , , , and see (43) and (45) below).
-
Case I. There is such that , or there is such that .
-
Case II. for all .
-
Case II.A. or or .
-
Case II.B. and and .
-
Case II.B.a. .
-
Case II.B.a.1. or .
-
Case II.B.a.1.A. or .
-
Case II.B.a.1.B. and .
-
-
-
-
Case II.B.a.2. and .
-
Case II. B.a.2.A. or is not constant on .
-
Case II. B.a.2.B. and .
-
-
-
Case II.B.b. .
Now let us discuss all the cases.
Proof of Case I. The proof is the same as that of Theorem 2.5, Case ℬ.I.
Proof of Case II. Suppose
for every . Notice
by (40) and (41).
Denote
Case II.A.
(a) Assume , i.e., for every , . Since is not constant on , so by Remark 2.4, , where . Thus, for every ,
Additionally, for all ,
Since is constant on , so it satisfies the conditions (i) and (ii) of Theorem 3.2 in [15]. Thus is a UM point, whence . Finally,
which finishes the proof.
(b) Assume . By (42), for every , . By (41) and ,
for all , whence . The rest of the proof goes as in (a).
(c) If , then analogous reasoning as in (b) goes for element .
Case II.B. Suppose and and . Clearly,
by (41) and
for all .
Case II.B.a. Assume . Then , since the opposite case (40) implies .
By Lemma 2.1 and Remark 2.1, we find the sets , , of measure satisfying
Clearly, by (44) and , the sets and are well defined.
Condition (41) implies
since otherwise if, for example, , then, by ,
a contradiction with (41).
Case II.B.a.1. Assume . Then implies
By (46), we have
Case II.B.a.1.A. Assume . Then by (47) and since . Thus
Moreover, by (40) and , we get .
We claim that
Assume for the contrary that . By Lemma 2.2, (48) and (49),
Moreover, by (46) and , we get . Furthermore, applying (41) and (45), we obtain
a contradiction with (51). This proves claim (50). Therefore, (48), (49) and (50) imply
Furthermore, by the definition of , for every ,
Thus
Finally,
Taking , we finish the proof (see Remark 2.3 and the definition of γ).
It is clear that analogous reasoning holds for the case of .
Case II.B.a.1.B. Assume and , whence
We claim that at least one of inequalities (56) or (57) holds,
If or , then by (47) and (55) we get (56) or (57), respectively.
If and , then
Assume for the contrary that (56) and (57) do not hold, i.e.,
By (46), we get
and
Therefore, by (59) and Lemma 2.2, we have
and
Thus, by (58), and , whence . Since , we get a contradiction with (40). This proves that (56) or (57) holds.
Therefore, by (48) and (56) or (57), we get . Analogously as in Case II.B.a.1.A, we get (53) and then (54), which for finishes the proof.
Considering the case of , we may follow analogously as above but with the element .
Case II.B.a.2. Suppose and . Then, by (46),
We claim and . If , then , whence by the definition of set , a contradiction with (42). The case of goes analogously and proves the claim.
By (39), and .
Case II.B.a.2.A. If is not constant on , then
Conditions (60) imply . Consequently,
Thus,
Finally, by (60), one of the following holds:
or
Analogously as in Case II.B.a.1.A, we get (53) and then (54), which for finishes the proof.
If is not constant on , then we use analogous argumentation.
Case II.B.a.2.B. Assume and . Note that this is only the case of since is not constant on .
Since for a.e. , and , so
for all .
If there is such that then, for every ,
If for every , then by the definition of we get (62) for every . Thus, by (61),
for .
Analogously as in Case II.B.a.1.A, we get (53) and then (54), which for finishes the proof.
Case II.B.b. If , then we apply the argumentation in Cases II.B.a.1 to II.B.a.2.A for γ instead of . □
The following corollaries have been proved directly in [20].
Corollary 2.1 The Lorentz space is nonsquare if and only if
-
(i)
,
-
(ii)
if then ,
-
(iii)
if then ,
where β and γ are defined in (6).
Proof Necessity. (i) Assume and take , where . Then . By Theorem 2.3, x is not an NSQ point.
(ii) and (iii) Suppose and , or and . Let . Then and Theorem 2.2 implies that x is not an NSQ point.
Sufficiency. Let .
Let . By Theorem 2.5, (i) and (iii), x is an NSQ point.
If then (ii) implies . In view of Theorem 2.4, x is an NSQ point. □
Recall that if and only if and (see [25]).
Corollary 2.2 Suppose and . The Lorentz space is nonsquare if and only if
-
(i)
,
-
(ii)
,
where β and γ are defined in (6).
Proof Necessity. (i) The proof is analogous to the proof of Corollary 2.1.
(ii) Let and take . Clearly, by Proposition 3.1 in [15], . The definition of x and Theorem 2.2 imply that x is not an NSQ point.
Sufficiency. Let . By Proposition 3.1 in [15], . Thus is not constant on . Moreover, (i) implies that . By Theorem 2.4, x is an NSQ point. □
References
García-Falset J, Llorens-Fuster E, Mazcuñan-Navarro EM: Uniformly nonsquare Banach spaces have the fixed point property for nonexpansive mappings. J. Funct. Anal. 2006, 233: 494–514. 10.1016/j.jfa.2005.09.002
James RC: Uniformly non-square Banach spaces. Ann. Math. 1964, 80: 542–550. 10.2307/1970663
James RC: Super-reflexive spaces with bases. Pac. J. Math. 1972, 41: 409–419. 10.2140/pjm.1972.41.409
Cerdà J, Hudzik H, Mastyło M: On the geometry of some Calderón-Lozanovskiĭ interpolation spaces. Indag. Math. 1995,6(1):35–49. 10.1016/0019-3577(95)98199-L
Cui YA, Jie L, Płuciennik R: Local uniform nonsquareness in Cesaro sequence spaces. Comment. Math. Prace Mat. 1997, 37: 47–58.
Denker M, Hudzik H:Uniformly non- Musielak-Orlicz sequence spaces. Proc. Indian Acad. Sci. Math. Sci. 1991, 101: 71–86. 10.1007/BF02868018
Foralewski P, Hudzik H, Kolwicz P: Non-squareness properties of Orlicz-Lorentz sequence spaces. J. Funct. Anal. 2013, 264: 605–629. 10.1016/j.jfa.2012.10.014
Foralewski P, Hudzik H, Kolwicz P: Non-squareness properties of Orlicz-Lorentz function spaces. J. Inequal. Appl. 2013. 10.1186/1029-242X-2013-32
Hudzik H:Uniformly non- Orlicz spaces with Luxemburg norm. Stud. Math. 1985, 81: 271–284.
Kamińska A, Kubiak D:On isometric copies of and James constants in Cesàro-Orlicz sequence spaces. J. Math. Anal. Appl. 2010, 372: 574–584. 10.1016/j.jmaa.2010.07.011
Kato M, Maligranda L: On James and Jordan-von Neumann constants of Lorentz sequence spaces. J. Math. Anal. Appl. 2001, 258: 457–465. 10.1006/jmaa.2000.7367
Kato M, Maligranda L, Takahashi Y: On James, Jordan-von Neumann constants and the normal structure coefficient of Banach spaces. Stud. Math. 2001, 144: 275–295. 10.4064/sm144-3-5
Maligranda L, Petrot N, Suantai S: On the James constant and B -convexity of Cesàro and Cesàro-Orlicz sequence spaces. J. Math. Anal. Appl. 2007, 326: 312–331. 10.1016/j.jmaa.2006.02.085
Beck A: A convexity condition in Banach spaces and the strong law of large numbers. Proc. Am. Math. Soc. 1962,13(2):329–334. 10.1090/S0002-9939-1962-0133857-9
Ciesielski M, Kolwicz P, Panfil A:Local monotonicity structure of Lorentz spaces . J. Math. Anal. Appl. 2014, 409: 649–662. 10.1016/j.jmaa.2013.07.028
Hudzik H, Kolwicz P, Narloch A: Local rotundity structure of Calderón-Lozanovskiĭ spaces. Indag. Math. 2006,17(3):373–395. 10.1016/S0019-3577(06)80039-X
Hudzik H, Narloch A: Local monotonicity structure of Calderón-Lozanovskiĭ spaces. Indag. Math. 2004,15(1):1–12. 10.1016/S0019-3577(04)90001-8
Kolwicz P, Płuciennik R:Local condition as a crucial tool for local structure of Calderón-Lozanovskiĭ spaces. J. Math. Anal. Appl. 2009, 356: 605–614. 10.1016/j.jmaa.2009.03.030
Kolwicz P, Płuciennik R: Points of upper local uniform monotonicity in Calderón-Lozanovskiĭ spaces. J. Convex Anal. 2010,17(1):111–130.
Kolwicz P, Panfil A:Non-square Lorentz spaces . Indag. Math. 2013, 24: 254–263. 10.1016/j.indag.2012.09.006
Lindenstrauss J, Tzafriri L: Classical Banach Spaces II. Springer, Berlin; 1979.
Bennett C, Sharpley R Pure and Applied Mathematics 129. In Interpolation of Operators. Academic Press, Boston; 1988.
Krein SG, Petunin JI, Semenov EM Transl. Math. Monogr. 54. In Interpolation of Linear Operators. Am. Math. Soc., Providence; 1982.
Ciesielski M, Kamińska A, Płuciennik R:Gâteaux derivatives and their applications to approximation in Lorentz spaces . Math. Nachr. 2009,282(9):1242–1264. 10.1002/mana.200810798
Kamińska A, Maligranda L:On Lorentz spaces . Isr. J. Math. 2004, 140: 285–318. 10.1007/BF02786637
Calderón AP: Intermediate spaces and interpolation, the complex method. Stud. Math. 1964, 24: 113–190.
Lorentz GG: On the theory of spaces Λ. Pac. J. Math. 1951, 1: 411–429. 10.2140/pjm.1951.1.411
Ariño MA, Muckenhoupt B: Maximal functions on classical Lorentz spaces and Hardy’s inequality with weights for nonincreasing functions. Trans. Am. Math. Soc. 1990,320(2):727–735.
Raynaud Y: On Lorentz-Sharpley spaces. Isr. Math. Conf. Proc. 1992, 5: 207–228.
Sawyer E: Boundedness of classical operators on classical Lorentz spaces. Stud. Math. 1990,96(2):145–158.
Ciesielski M, Kamińska A, Kolwicz P, Płuciennik R:Monotonicity and rotundity properties of Lorentz spaces . Nonlinear Anal. 2012, 75: 2713–2723. 10.1016/j.na.2011.11.011
Acknowledgements
The first author (Paweł Kolwicz) is supported by the Ministry of Science and Higher Education of Poland, grant number 04/43/DSPB/0083 (including the article processing charge).
Author information
Authors and Affiliations
Corresponding author
Additional information
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
Both authors have made intellectual contributions to a published study in equal parts and have written the manuscript. Authors read and approved the final manuscript.
Rights and permissions
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0), which permits use, duplication, adaptation, distribution, and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
About this article
Cite this article
Kolwicz, P., Panfil, A. Points of nonsquareness of Lorentz spaces . J Inequal Appl 2014, 467 (2014). https://doi.org/10.1186/1029-242X-2014-467
Received:
Accepted:
Published:
DOI: https://doi.org/10.1186/1029-242X-2014-467