The index of strong rotundity is introduced. This index is used to determine how far an element of the unit sphere of a real Banach space is from being a strongly exposed point of the unit ball. This index is computed for Hilbert spaces. Characterizations of the set of rotund points and the set of smooth points are provided for a better understanding of the construction of the index of strong rotundity. Finally, applications to the stereographic projection are provided.
Citation: Francisco Javier García-Pacheco. The index of strong rotundity[J]. AIMS Mathematics, 2023, 8(9): 20477-20486. doi: 10.3934/math.20231043
[1] | Almudena Campos-Jiménez, Francisco Javier García-Pacheco . The core of the unit sphere of a Banach space. AIMS Mathematics, 2024, 9(2): 3440-3452. doi: 10.3934/math.2024169 |
[2] | Sani Salisu, Vasile Berinde, Songpon Sriwongsa, Poom Kumam . Approximating fixed points of demicontractive mappings in metric spaces by geodesic averaged perturbation techniques. AIMS Mathematics, 2023, 8(12): 28582-28600. doi: 10.3934/math.20231463 |
[3] | Junaid Ahmad, Kifayat Ullah, Reny George . Numerical algorithms for solutions of nonlinear problems in some distance spaces. AIMS Mathematics, 2023, 8(4): 8460-8477. doi: 10.3934/math.2023426 |
[4] | Hsien-Chung Wu . Near fixed point theorems in near Banach spaces. AIMS Mathematics, 2023, 8(1): 1269-1303. doi: 10.3934/math.2023064 |
[5] | Wenlong Sun, Gang Lu, Yuanfeng Jin, Choonkil Park . Self-adaptive algorithms for the split problem of the quasi-pseudocontractive operators in Hilbert spaces. AIMS Mathematics, 2022, 7(5): 8715-8732. doi: 10.3934/math.2022487 |
[6] | Meiying Wang, Luoyi Shi, Cuijuan Guo . An inertial iterative method for solving split equality problem in Banach spaces. AIMS Mathematics, 2022, 7(10): 17628-17646. doi: 10.3934/math.2022971 |
[7] | Lu-Chuan Ceng, Yeong-Cheng Liou, Tzu-Chien Yin . On Mann-type accelerated projection methods for pseudomonotone variational inequalities and common fixed points in Banach spaces. AIMS Mathematics, 2023, 8(9): 21138-21160. doi: 10.3934/math.20231077 |
[8] | Gul Rahmat, Tariq Shah, Muhammad Sarwar, Saber Mansour, Hassen Aydi . Strong convergence to fixed points of an evolution subfamily. AIMS Mathematics, 2023, 8(9): 20380-20394. doi: 10.3934/math.20231039 |
[9] | Xun Ge, Songlin Yang . Some fixed point results on generalized metric spaces. AIMS Mathematics, 2021, 6(2): 1769-1780. doi: 10.3934/math.2021106 |
[10] | Kifayat Ullah, Junaid Ahmad, Hasanen A. Hammad, Reny George . Iterative schemes for numerical reckoning of fixed points of new nonexpansive mappings with an application. AIMS Mathematics, 2023, 8(5): 10711-10727. doi: 10.3934/math.2023543 |
The index of strong rotundity is introduced. This index is used to determine how far an element of the unit sphere of a real Banach space is from being a strongly exposed point of the unit ball. This index is computed for Hilbert spaces. Characterizations of the set of rotund points and the set of smooth points are provided for a better understanding of the construction of the index of strong rotundity. Finally, applications to the stereographic projection are provided.
The irruption of the moduli of convexity and smoothness [7] in the literature of the geometry of the real Banach spaces was a huge revolution that brought strong implications to longstanding open problems such as the Banach-Mazur conjecture for rotations or the fixed-point property problem. The main purpose of this manuscript is to introduce an index that measures a convexity stronger than strict convexity but weaker than uniform convexity. All Banach spaces considered throughout this manuscript will be over the reals. A point x in the unit sphere SX of a Banach space X is said to be a strongly exposed point of the unit ball BX if there exists x∗ in the unit sphere SX∗ of the dual space X∗ verifying the following property: If (xn)n∈N⊆BX is such that (x∗(xn))n∈N converges to 1, then (xn)n∈N converges to x. The functional x∗ is said to strongly expose x on BX and it is trivial that (x,x∗)∈ΠX, where ΠX:={(x,x∗)∈SX×SX∗:x∗(x)=1}. We will let
ΠseX:={(x,x∗)∈SX×SX∗:x∗ strongly exposes x on BX}. |
A weaker notion than a strongly exposed point is that of an exposed point. A point x∈SX is said to be an exposed point of BX if there exists x∗∈SX∗ in such a way that (x∗)−1({1})∩BX={x}. This time the functional x∗ is called a supporting functional that exposes x on BX. We will let ΠeX:={(x,x∗)∈SX×SX∗:x∗ exposes x on BX}. Observe that ΠseX⊆ΠeX⊆ΠX. Another trivial fact is the following: for every surjective linear isometry T between Banach spaces X,Y and for every (x,x∗)∈SX×SX∗, we have that (x,x∗)∈ΠX or ΠeX or ΠseX if and only if (T(x),T∗(x∗))∈ΠY or ΠeY or ΠseY, respectively. In other words, the previous notions are invariant under surjective linear isometries. Another geometrical notion employed in this manuscript is that of a rotund point. A point x∈SX in the unit sphere of a Banach space X is said to be a rotund point of the unit ball of X if x is contained in no non-trivial segment of the unit sphere, in other words, {x} is a maximal proper face of BX. The set of rotund points of BX is denoted by rot(BX). In view of the Hahn-Banach Separation Theorem, the set of rotund points can be described as follows:
rot(BX)={x∈SX:if x∗∈SX∗ is so that (x,x∗)∈ΠX, then (x,x∗)∈ΠeX}. |
We refer the reader to [2,3] for a wider perspective on the above concepts and some other geometrical properties related with renormings. The duality mapping [4,5] of a Banach space X is the set-valued map defined as
J:X→P(X∗)x↦J(x):={x∗∈X∗:‖x∗‖=‖x‖ and x∗(x)=‖x∗‖‖x‖}. |
A point x in the unit sphere SX of X is said to be a smooth point [8] of the unit ball BX of X provided that J(x) is a singleton. The subset of smooth points of BX is typically denoted by smo(BX). Rotund points and smooth points are dual notions.
The main goal of this manuscript is to introduce an index that measures accurately how far a point in the unit sphere is from being a strongly exposed point of the unit ball. For this, we first establish characterizations of the set of rotund points and the set of smooth points. Then we introduce the index of strong rotundity and show the most basic properties related to such index. We compute the index of strong rotundity of a Hilbert space. Finally, we construct a new set of pairs contained in ΠeX for which the stereographic projection [13,14] in a Banach space is a homeomorphism.
We will begin by providing a new characterization of the set of rotund points of the unit ball of a Banach space. As usual, if X is a vector space and x,y∈X, then st(x,y):=x+R(y−x) is the straight line passing through x,y and [x,y]:=x+[0,1](y−x) is the segment joining x,y.
Theorem 2.1. For a Banach space X, rot(BX)={x∈SX:∀y∈SXst(x,y)∩SX={x,y}}.
Proof. We will prove both inclusions in two simple steps.
⊆ Let x∈rot(BX). Fix an arbitrary y∈SX∖{x}. If there exists z∈st(x,y)∩SX different from x,y, then we end up with three points aligned in the unit sphere, so the whole segment containing x,y,z lies entirely in the unit sphere. However, this contradicts the fact that {x} is a maximal proper face of BX.
⊇ Conversely, let x∈SX satisfying that st(x,y)∩SX={x,y} for all y∈SX. If x∉rot(BX), then we can find y∈SX∖{x} such that [x,y]⊆SX, contradicting that st(x,y)∩SX={x,y} since [x,y]⊆SX.
The next result is a characterization of the set of smooth points in similar terms as the previous theorem. However, a technical lemma is needed first.
Lemma 2.1. Let X be a Banach space. Let x∈SX and consider a straight line L⊆X containing x such that L∩(SX∖{x})=∅. Then L∩UX=∅, where UX is the open unit ball of X.
Proof. Suppose on the contrary that there exists u∈L∩UX. Consider the continuous function
(−∞,0]→Rt↦‖u+t(x−u)‖. |
If t=0, then ‖u‖<1. Note that
1+‖u‖‖x−u‖≥1>‖u‖, |
therefore, if t<−1+‖u‖‖x−u‖, then
‖u+t(x−u)‖≥|‖u‖−‖t(x−u)‖|=|‖u‖−|t|‖x−u‖|=|t|‖x−u‖−‖u‖>1. |
Bolzano's Theorem assures the existence of s∈(−∞,0) such that ‖u+s(x−u)‖=1, that is, u+s(x−u)∈L∩SX, meaning that u+s(x−u)=x, hence (1−s)u=(1−s)x, so u=x. This is a contradiction because ‖u‖<1=‖x‖.
Theorem 2.2. For a Banach space X with dim(X)≥2,
smo(BX)={x∈SX:∃x∗∈SX∗(x,x∗)∈ΠX and ∀z∈(x∗)−1({1})st(−x,z)∩(SX∖{−x})≠∅}. |
Proof. We will prove both inclusions in two steps.
⊆ Let x∈smo(BX). There exists x∗∈SX∗ satisfying that J(x)={x∗}. Fix an arbitrary z∈(x∗)−1({1}). Suppose on the contrary that st(−x,z)∩(SX∖{−x})=∅. Note that, in this case, ‖z‖>1 because if ‖z‖=1, then z∈st(−x,z)∩(SX∖{−x}). By bearing in mind Lemma 2.1, st(−x,z)∩UX=∅. The Hahn-Banach Separation Theorem allows the existence of y∗∈SX∗ such that st(−x,z)⊆(y∗)−1({1}). Then y∗(−x)=1, meaning that −y∗(x)=1, so −y∗∈J(x)={x∗}, reaching the contradiction that y∗(z)=1 and y∗(z)=−x∗(z)=−1.
⊇ Take x∈SX for which there exists x∗∈J(x) satisfying that st(−x,z)∩(SX∖{−x})≠∅ for all z∈(x∗)−1({1}). Suppose on the contrary that x∉smo(BX). There exists a 2-dimensional subspace Y of X containing x for which x is not a smooth point of BY. Let JY:Y→P(Y∗) denote the dual mapping of Y. Notice that JY(x) is a nontrivial segment that lies entirely in SY∗. Thus we can write JY(x)=[a∗,b∗] where a∗≠b∗ both are in SY∗. Notice that BY⊆(x∗|Y)−1([−1,1])∩(a∗)−1([−1,1])∩(b∗)−1([−1,1]). Also, x∗|Y∈JY(x)=[a∗,b∗], therefore, either a∗≠x∗|Y or b∗≠x∗|Y. Let us assume without any loss of generality that a∗≠x∗|Y. Then (a∗)−1({−1}) is not parallel to (x∗|Y)−1({1}), hence the straight line (a∗)−1({−1}) intersects (x∗|Y)−1({1}). Choose any z∈(x∗|Y)−1({1}) with (a∗)(z)<−1. On the one hand, st(−x,z)⊆Y, thus st(−x,z)∩(SX∖{−x})⊆st(−x,z)∩(SY∖{−x}). On the other hand, we will prove that st(−x,z)∩(SY∖{−x})=∅. Indeed, pick any t∈R∖{0} and distinguish the following two cases:
● t>0. In this case, a∗(tz+(1−t)(−x))=t(a∗(z)+1)−1<−1. Since BY⊆(a∗)−1([−1,1]), we conclude that tz+(1−t)(−x)∉BY.
● t<0. In this case, x∗|Y(tz+(1−t)(−x))=2t−1<−1. Since BY⊆(x∗|Y)−1([−1,1]), we conclude that tz+(1−t)(−x)∉BY.
As a consequence, we end up having the contradiction that st(−x,z)∩(SX∖{−x})=∅.
Let us introduce next the index of strong rotundity of a Banach space. First, we recall that, for a Banach space X, UX stands for the open unit ball, and UX(x,ε) denotes the open ball of center x∈X and radius ε≥0.
Definition 2.1. (Index of strong rotundity) Let X be a Banach space. The local index of strong rotundity of X at (x,x∗)∈ΠX is defined as
ηX(⋅,(x,x∗)):[0,2]→[0,2]ε↦ηX(ε,(x,x∗)):=d((x∗)−1({1}),BX∖UX(x,ε)). | (2.1) |
The index of strong rotundity of X at is defined as
ηX:[0,2]→[0,2]ε↦ηX(ε):=inf{ηX(ε,(x,x∗)):(x,x∗)∈ΠX}. | (2.2) |
The index of strong rotundity will serve, among other things, to characterize the strongly exposed points of the unit ball. Later on, we will relate the index of strong rotundity to the modulus of local uniform rotundity. Now, the following lemma unveils the most basic properties satisfied by the index of rotundity.
Lemma 2.2. Let X be a Banach space. Let (x,x∗)∈ΠX. Then:
(1) ηX(0,(x,x∗))=0.
(2) (x∗)−1({−1})∩BX⊆BX∖UX(x,2), hence ηX(2,(x,x∗))≤2.
(3) If (x∗)−1({−1})∩BX=BX∖UX(x,2), then ηX(2,(x,x∗))=2.
(4) If 0≤ε1≤ε2≤2, then ηX(ε1,(x,x∗))≤ηX(ε2,(x,x∗)).
(5) If 0≤ε1≤ε2≤2, then ηX(ε1)≤ηX(ε2).
(6) ΠseX={(x,x∗)∈ΠX:∀ε∈(0,2]ηX(ε,(x,x∗))>0}.
Proof. We will only prove the second, third, and last items.
(2) For every z∈(x∗)−1({−1})∩BX, 2≥‖z−x‖≥|x∗(z−x)|=|x∗(z)−x∗(x)|=2, meaning that z∈BX∖UX(x,2).
(3) On the one hand, if u∈(x∗)−1({1}) and v∈(x∗)−1({−1})∩BX, then ‖u−v‖≥|x∗(u−v)|=|x∗(u)−x∗(v)|=2, thus d((x∗)−1({1}),(x∗)−1({−1})∩BX)=2. On the other hand,
2≥ηX(2,(x,x∗)):=d((x∗)−1({1}),BX∖UX(x,ε))=d((x∗)−1({1}),(x∗)−1({−1})∩BX)=2. |
(6) Fix an arbitrary (x,x∗)∈ΠseX. Suppose on the contrary that there exists ε∈(0,2] for which ηX(ε,(x,x∗))=0. Then we can find two sequences (yn)n∈N⊆(x∗)−1({1}) and (xn)n∈N⊆BX∖UX(x,ε) such that ‖yn−xn‖→0 as n→∞. Notice that x∗(xn)→1 as n→∞ but (xn)n∈N does not converge to x because xn∉UX(x,ε) for each n∈N, contradicting that (x,x∗)∈ΠseX. Conversely, take any (x,x∗)∈ΠX satisfying that ηX(ε,(x,x∗))>0 for all ε∈(0,2]. Assume to the contrary that (x,x∗)∉ΠseX. There exists a sequence (xn)n∈N⊆BX in such a way that x∗(xn)→1 as n→∞ but (xn)n∈N does not converge to x. By passing to an appropriate subsequence (xnk)k∈N, we can find ε∈(0,2] such that ‖xnk−x‖≥ε for all k∈N. Notice that (x∗(xnk))k∈N still converges to 1, so, by assuming that x∗(xnk)≠0 for all k∈N, we have that ‖xnkx∗(xnk)−xnk‖→0, meaning that ηX(ε,(x,x∗))=d((x∗)−1({1}),BX∖UX(x,ε))=0, which contradicts our initial assumption.
Notice that, under the settings of Lemma 2.2, it does not always hold that (x∗)−1({−1})∩BX=BX∖UX(x,2).
Proposition 2.1. Let X be a Banach space. Let (x,x∗)∈ΠeX. If x∉rot(BX), then (x∗)−1({−1})∩BX⊊BX∖UX(x,2).
Proof. In accordance with Lemma 2.2(2), (x∗)−1({−1})∩BX⊆BX∖UX(x,2). Since (x,x∗)∈ΠeX, we have that (x∗)−1({1})∩BX={x}, so (x∗)−1({−1})∩BX={−x}. Since x∉rot(BX), there exists a non-trivial segment of the unit sphere containing x, that is, there exists y∈SX∖{x} with ‖x+y2‖=1, in other words, ‖x+y‖=2, meaning that −y∈BX∖UX(x,2). Finally, −y∉{−x}=(x∗)−1({−1})∩BX. As a consequence, (x∗)−1({−1})∩BX⊊BX∖UX(x,2).
In ℓ∞, we will let 1 to denote the constant sequence of general term 1. Also, (en)n∈N will denote the sequence of canonical unit vectors, that is, en(m)=δnm for all n,m∈N. And
δn:ℓ∞→Rx↦δn(x):=x(n) | (2.3) |
is the nth-coordinate functional. Observe that (1,δ1),(e1,δ1)∈Πℓ∞.
Proposition 2.2. ηℓ∞(ε,(1,δ1))=0 for all ε∈[0,2] and ηℓ∞(2,(e1,δ1))=2.
Proof. According to Lemma 2.2(4), it only suffices to prove that ηℓ∞(2,(1,δ1))=0. For this, we will show that (δ1)−1({1})∩(Bℓ∞∖Uℓ∞(1,2))≠∅. Indeed,
(1,−1,−1,−1,…)∈(δ1)−1({1})∩(Bℓ∞∖Uℓ∞(1,2)). |
On the other hand, Bℓ∞∖Uℓ∞(e1,2)=(δ1)−1({−1})∩Bℓ∞, thus
ηℓ∞(2,(e1,δ1))=d((δ1)−1({1}),Bℓ∞∖Uℓ∞(e1,2))=d((δ1)−1({1}),(δ1)−1({−1})∩Bℓ∞)=2 |
in view of Lemma 2.2(3).
Recall [1,6] that, given a Banach space X with unit sphere SX and unit ball BX, a closed subspace Y⊆X is said to be an Lp-summand subspace of X, where 1≤p≤∞, if Y is Lp-complemented in X, that is, there exists a closed subspace Z⊆X such that X=Y⊕pZ, in the sense that ‖y+z‖p=‖y‖p+‖z‖p for all y∈Y and all z∈Z. A point x∈X is said to be an Lp-summand vector of X provided that Rx is an Lp-summand subspace of X. In accordance with [1], every unit L2-summand vector is a rotund point as well as a smooth point of the unit ball.
Theorem 2.3. Let X be a Banach space with dim(X)≥2. For every (x,x∗)∈ΠX such that x is an L2-summand vector of X and for every ε∈[0,2],
ηX(ε,(x,x∗))=ε22. |
Proof. First off, notice that ker(x∗) is the L2-complement of Rx. For every z∈(x∗)−1({1}) and every y∈BX∖UX(x,ε), ‖z−y‖≥|x∗(z−y)|=|x∗(z)−x∗(y)|=1−x∗(y). Fix an arbitrary y∈BX∖UX(x,ε) and write y=x∗(y)x+my with my∈ker(x∗). Notice that x∗(y)2+‖my‖2≤1 and (1−x∗(y))2+‖my‖2=‖x−y‖2≥ε2. Then (1−x∗(y))2≥ε2−‖my‖2≥ε2+x∗(y)2−1. In other words, (1−x∗(y))2+1−x∗(y)2≥ε2, that is, 2−2x∗(y)≥ε2, meaning that 1−x∗(y)≥ε22. Going back to the beginning, for every z∈(x∗)−1({1}) and every y∈BX∖UX(x,ε), ‖z−y‖≥1−x∗(y)≥ε22, which implies that ηX(ε,(x,x∗))≥ε22. Finally, since dim(X)≥2, we can take any y∈SX∩SX(x,ε) and z:=x+my, where my:=y−x∗(y)x. Simply notice that ‖z−y‖=‖(x+my)−y‖=‖(x+y−x∗(y)x)−y‖=‖x−x∗(y)x‖=1−x∗(y)=ε22.
Corollary 2.1. If X is a Hilbert space with dim(X)≥2, then ηX(ε)=ε22.
Proof. By bearing in mind [6], a Banach space is a Hilbert space if and only every unit vector is an L2-summand vector. Therefore, by applying Theorem 2.3, ηX(ε,(x,x∗))=ε22 for every (x,x∗)∈ΠX, obtaining the desired result.
Another index can be defined which lies in between the local modulus of convexity and the index of strong rotundity. Indeed, if X is a Banach space, then we define at (x,x∗)∈ΠX the following function:
υX(⋅,(x,x∗)):[0,2]→[0,2]ε↦υX(ε,(x,x∗)):=inf{1−x∗(y):‖y‖≤1,‖x−y‖≥ε)}. | (2.4) |
We remind the reader that the modulus of local convexity [12] at x∈SX is given by
δX(⋅,x):[0,2]→[0,1]ε↦δX(ε,x):=inf{1−‖x+y2‖:‖y‖≤1,‖x−y‖≥ε)}. | (2.5) |
Theorem 2.4. Let X be a Banach space. For every (x,x∗)∈ΠX and every ε∈[0,2],
2δX(ε,x)≤υX(ε,(x,x∗))≤ηX(ε,(x,x∗)). |
Proof. On the one hand, for every z∈(x∗)−1({1}) and every y∈BX∖UX(x,ε),
‖z−y‖≥x∗(z−y)=1−x∗(y)≥υX(ε,(x,x∗)). |
As a consequence, ηX(ε,(x,x∗))≥υX(ε,(x,x∗)). On the other hand, for every y∈BX∖UX(x,ε),
δX(ε,x)≤1−‖x+y2‖≤1−x∗(x+y2)=1−1+x∗(y)2=1−x∗(y)2. |
Therefore, δX(ε,x)≤12υX(ε,(x,x∗)).
Let us finally tackle some applications to the stereographic projection. According to [9,10], if X is a Banach space and (x,x∗)∈ΠeX, then
SX∖{−x}→(x∗)−1({1})y↦−x+2x∗(y)+1(y+x) | (2.6) |
is a well-defined and continuous function known as stereographic projection.
Definition 2.2. (Stereographic projection pair) Let X be a Banach space. Let (x,x∗)∈ΠX. We will say that (x,x∗) is a stereographic projection pair provided that the following conditions are satisfied:
● st(−x,y)∩SX={−x,y} for all y∈SX.
● st(−x,z)∩(SX∖{−x})≠∅ for every z∈(x∗)−1({1}).
● If (yj)j∈N⊆SX∖{−x} is a sequence converging to −x, then ‖yj+xx∗(yj)+1‖→∞ as j→∞.
The set of stereographic projection pairs will be denoted by ΠspX.
According to Theorem 2.1, the first condition in the above definition is equivalent to the fact that −x∈rot(BX). The second condition is equivalent to the fact that x∈smo(BX) in view of Theorem 2.2. As a consequence, if (x,x∗)∈ΠspX, then x∈rot(BX)∩smo(BX), hence (x,x∗)∈ΠeX. Nevertheless, observe that the last condition of the previous definition is an unusual geometrical property in the sense that it only works for sequences in the unit sphere, but not for sequences in the unit ball. Indeed, if we take yj:=tjx for all j∈N, where (tj)j∈N⊆(−1,1) converges to −1, then
‖yj+xx∗(yj)+1‖=‖tjx+xtj+1‖=1 |
for all j∈N.
By bearing in mind [9, Lemma 2.1], if x∈SX is an L2-summand vector of a Banach space X and x∗∈SX∗ satisfies that ker(x∗) is the L2-complement of Rx, then (x,x∗) is a stereographic projection pair. Since a Banach space is a Hilbert space if and only every unit vector is an L2-summand vector [6], we obtain the following theorem whose proof we omit.
Theorem 2.5. If X is a Hilbert space, then ΠX=ΠspX.
The following theorem generalizes and improves [9, Lemma 2.1]. First, a technical remark is needed, which is a simple limit from a Calculus course.
Remark 2.1. For each p∈(1,∞), limx→−1+1−(−x)p(1+x)p=+∞.
Theorem 2.6. Let X be a Banach space with dim(X)≥2. Let (x,x∗)∈ΠX. If x is an Lp-summand vector of X, for 1<p<∞, and ker(x∗) is the Lp-complement of Rx, then (x,x∗) is a stereographic projection pair.
Proof. In the first place, x is an Lp-summand vector of any 2-dimensional subspace containing it. Therefore, any 2-dimensional subspace containing x is linearly isometric to ℓ2p, which is rotund and smooth. Since rotund points and smooth points are 2-dimensional properties, we conclude that x∈rot(BX)∩smo(BX). By applying Theorem 2.1 and Theorem 2.2, we conclude that the first two conditions of Definition 2.2 are satisfied. Let us prove the third condition. Take any sequence (yj)j∈N⊆SX∖{−x} converging to −x. Let us write yj=x∗(yj)x+mj, where mj∈ker(x∗). Since (x∗(yj))j∈N converges to −1, we may assume that −1<x∗(yj)<0 for all j∈N. Notice then that 1=(−x∗(yj))p+‖mj‖p for all j∈N. Then
‖yj+xx∗(yj)+1‖p=(1+x∗(yj))p+‖mj‖p(x∗(yj)+1)p=(1+x∗(yj))p+1−(−x∗(yj))p(x∗(yj)+1)p=1+1−(−x∗(yj))p(x∗(yj)+1)p→∞ |
as j→∞ in view of Remark 2.1.
The final result in this manuscript shows that stereographic projection pairs make possible that stereographic projections in Banach spaces be homeomorphisms, improving [9, Theorem 2.2]. However, let us recall first the following well-known topological fact [11].
Remark 2.2. Let X,Y be topological spaces. Let f:X→Y be injective. Let x∈X. Suppose that for every net (xi)i∈I⊆X such that (f(xi))i∈I converges to f(x), there exists a subnet (zj)j∈J of (xi)i∈I convergent to x. Then f−1:f(X)→X is continuous at f(x).
One can easily understand that, under the settings of Remark 2.2, if X,Y are both first countable, then Remark 2.2 remains true if we switch nets with sequences.
Theorem 2.7. Let X be a Banach space. If (x,x∗)∈ΠspX, then the stereographic projection (2.6) is an homeomorphism.
Proof. First off, let us denote by ϕ to the stereographic projection (2.6). We already know that ϕ is well defined, continuous, and ϕ(y)∈st(−x,y) for all y∈SX∖{−x}. Let us check now that ϕ is surjective. Fix an arbitrary z∈(x∗)−1({1}). If z=x, then ϕ(x)=x. So let us assume that z≠x. Since (x,x∗)∈ΠspX, by definition we have that st(−x,z)∩(SX∖{−x})≠∅. Let u∈R∖{0} such that y:=−x+u(z+x)∈SX∖{−x}. We will show that ϕ(y)=z. Indeed
ϕ(y)=−x+2y+xx∗(y)+1=−x+2−x+u(z+x)+xx∗(−x+u(z+x))+1=−x+2u(z+x)2u=−x+(z+x)=z. |
Next step is to prove that ϕ is one-to-one. Indeed, take y1,y2∈SX∖{−x} with ϕ(y1)=ϕ(y2). Then y2=−x+x∗(y2)+1x∗(y1)+1(y1+x)∈st(−x,y1)∩SX={−x,y1}, meaning that y1=y2. Let us finally prove that ϕ−1 is continuous. We will rely on Remark 2.2 for sequences. Fix an arbitrary y∈SX∖{−x}. Take a sequence (yi)i∈N⊆SX∖{−x} such that (ϕ(yi))i∈N converges to ϕ(y). We will show the existence of a subsequence (yij)j∈N convergent to y. Indeed, there exists a subsequence (yij)j∈N such that (x∗(yij))j∈N is convergent to some r∈[−1,1]. Then (ϕ(yij))j∈N converges to ϕ(y). This is equivalent to saying that (yij+xx∗(yij)+1)j∈N converges to y+xx∗(y)+1. Since (x∗(yij)+1)j∈N is convergent to r+1, we conclude that ((x∗(yij)+1)yij+xx∗(yij)+1)j∈N converges to (r+1)y+xx∗(y)+1, in other words, (yij+x)j∈N converges to (r+1)y+xx∗(y)+1, which is equivalent to stating that (yij)j∈N converges to −x+(r+1)y+xx∗(y)+1. Next, observe that r≠−1 since otherwise we obtain that (yij)j∈N converges to −x, reaching the contradiction that ‖yij+xx∗(yij)+1‖→∞ as j→∞ by bearing in mind that (x,x∗)∈ΠspX. As a consequence, −x+(r+1)y+xx∗(y)+1∈st(−x,y)∩SX={−x,y}, that is, either −x+(r+1)y+xx∗(y)+1=−x or −x+(r+1)y+xx∗(y)+1=y. If −x+(r+1)y+xx∗(y)+1=−x, then y=−x, which is impossible since y∈SX∖{−x}. Thus, −x+(r+1)y+xx∗(y)+1=y. By relying on Remark 2.2, we conclude that ϕ−1 is continuous at b.
The use of indices in the literature of Geometry of Banach Spaces has always been very useful to determine the exact shape of the unit ball. The most known indices are the modulus of convexity, the modulus of smoothness, the index of rotundity, and the Bishop-Phelps-Bollabás index, among others. The index of strong rotundity is a novel concept introduced in this manuscript. This index serves to determine how far a point of the unit sphere is from being an strongly exposed point. As expected, Hilbert spaces have a very particular index of strong rotundity. Finally, applications to the stereographic projection are provided, in particular, a novel set of pairs (the stereographic projection pairs) are introduced in this manuscript which guarantee that the stereographic projection is an homeomorphism.
The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.
This research was funded by Ministerio de Ciencia, Innovación y Universidades: PGC-101514-B-I00 (Métodos analíticos en Simetrías, Teoría de Control y Operadores); and Consejería de Universidad, Investigación e Innovación de la Junta de Andalucía: FEDER-UCA18-105867 (Dispositivos electrónicos para la estimulación magnética transcraneal), ProyExcel00780 (Operator Theory: An interdisciplinary approach), and ProyExcel01036 (Multifísica y optimización multiobjetivo de estimulación magnética transcraneal).
The author declares that there is no conflict of interest.
[1] |
A. Aizpuru, F. J. García-Pacheco, L2-summand vectors in Banach spaces, Proc. Amer. Math. Soc., 134 (2006), 2109–2115. https://doi.org/10.1090/S0002-9939-06-08243-8 doi: 10.1090/S0002-9939-06-08243-8
![]() |
[2] |
P. Bandyopadhyay, D. Huang, B. L. Lin, S. L. Troyanski, Some generalizations of locally uniform rotundity, J. Math. Anal. Appl., 252 (2000), 906–916. https://doi.org/10.1006/jmaa.2000.7169 doi: 10.1006/jmaa.2000.7169
![]() |
[3] |
P. Bandyopadhyay, B. L. Lin, Some properties related to nested sequence of balls in {B}anach spaces, Taiwanese J. Math., 5 (2001), 19–34. https://doi.org/10.11650/twjm/1500574887 doi: 10.11650/twjm/1500574887
![]() |
[4] |
A. Beurling, A. E. Livingston, A theorem on duality mappings in Banach spaces, Ark. Mat., 4 (1962), 405–411. https://doi.org/10.1007/BF02591622 doi: 10.1007/BF02591622
![]() |
[5] | J. Blažek, Some remarks on the duality mapping, Acta Univ. Carolin. Math. Phys., 23 (1982), 15–19. |
[6] | J. W. Carlson, T. L. Hicks, A characterization of inner product spaces, Math. Japon, 23 (1978), 371–373. |
[7] |
J. A. Clarkson, Uniformly convex spaces, Trans. Amer. Math. Soc., 40 (1936), 396–414. https://doi.org/10.2307/1989630 doi: 10.2307/1989630
![]() |
[8] |
D. F. Cudia, The geometry of Banach spaces. Smoothness, Trans. Amer. Math. Soc., 110 (1964), 284–314. https://doi.org/10.2307/1993705 doi: 10.2307/1993705
![]() |
[9] |
F. J. García-Pacheco, J. B. Seoane-Sepúlveda, The stereographic projection in Banach spaces, Rocky Mountain J. Math., 37 (2007), 1167–1172. https://doi.org/10.1216/rmjm/1187453103 doi: 10.1216/rmjm/1187453103
![]() |
[10] |
F. J. García-Pacheco, The stereographic projection in topological modules, Axioms, 12 (2023), 225. https://doi.org/10.3390/axioms12020225 doi: 10.3390/axioms12020225
![]() |
[11] | J. L. Kelley, General topology, New York: Springer-Verlag, 1975. |
[12] |
A. R. Lovaglia, Locally uniformly convex Banach spaces, Trans. Amer. Math. Soc., 78 (1955), 225–238. https://doi.org/10.2307/1992956 doi: 10.2307/1992956
![]() |
[13] |
C. F. Marcus, The stereographic projection in vector notation, Math. Mag., 39 (1966), 100–102. https://doi.org/10.1080/0025570X.1966.11975692 doi: 10.1080/0025570X.1966.11975692
![]() |
[14] |
A. P. Stone, On the stereographic projection of the sphere, Math. Gaz., 40 (1956), 181–184, https://doi.org/10.2307/3608806 doi: 10.2307/3608806
![]() |
1. | Almudena Campos-Jiménez, Francisco Javier García-Pacheco, The core of the unit sphere of a Banach space, 2024, 9, 2473-6988, 3440, 10.3934/math.2024169 |