The distance function and Lipschitz classes of mappings between metric spaces

Marijan Marković · 2025

Read The distance function and Lipschitz classes of mappings between metric spaces by Marijan Marković (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-14508.txt.

Introduction

Marković, writing solo in twenty twenty-five, asked when a local smoothness condition forces a mapping between metric spaces to be globally Lipschitz. Lipschitz mappings stretch distances by at most a fixed factor. He studies the distance from a point's image to a fixed set, a real-valued function built from the mapping. When that function is locally Lipschitz, the mapping itself is globally Lipschitz. As a consequence, an analytic function belongs to the Lipschitz class exactly when its modulus does, extending a classical disk result.

Catalog record

CREATORMarijan Marković
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE29 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-14508

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

The distance function and Lipschitz classes of mappings between metric spaces Thanks: This research was supported by the Ministry of Education, Science, and Innovation of Montenegro through the grant ”Mathematical Analysis, Optimization, and Machine Learning”. Marijan Marković Abstract . We investigate when the local Lipschitz property of the real-valued function g ⁡ ( z ) = d Y ​ ( f ⁡ ( z ) , A ) g(z)=d_{Y}(f(z),A) implies the global Lipschitz property of the mapping f : X → Y f:X\to Y between the metric spaces ( X , d X ) (X,d_{X}) and ( Y , d Y ) (Y,d_{Y}) . Here, d Y ​ ( y , A ) d_{Y}(y,A) denotes the distance of y ∈ Y y\in Y from the non-empty set ⊆ Y \subseteq Y . As a consequence, we find that an analytic function on a uniform domain of a normed space belongs to the Lipschitz class if and only if its modulus satisfies the same condition; in the case of the unit disk this result is proved by K. Dyakonov. We use the recently established version of a classical theorem by Hardy and Littlewood for mappings between metric spaces. This paper is a continuation of the recent article by the author [ 14 ] . 2020 Mathematics Subject Classification Primary 26A16, 46E50; Secondary 51F30, 46G20, 30L15 1. Introduction This paper is mainly motivated by a result for analytic functions on the unit disk \mathbb{D}\subseteq\mathbb{C} obtained by Dyakonov [ 4 ] . Among others, he deduced the following interesting result. An analytic function belongs to the certain Lipschitz class if and only if its modulus belongs to the same Lipschitz class. In a special case, for \alpha\in(0,1) we have the following two equivalent statements: f\in\Lambda^{\alpha}(\mathbb{D})\Leftrightarrow|f|\in\Lambda^{\alpha}(\mathbb{D}). We have denoted by \Lambda^{\alpha}(\mathbb{D})=\{f:\mathbb{D}\to\mathbb{C}:\exists C_{f}\text{ such that }|f(z)-f(w)|\leq C_{f}|z-w|^{\alpha}\text{ for }z,w\in\mathbb{D}\} the Lipschitz (Hölder) class of functions on the unit disk. Pavlović [ 15 ] gave a new proof of the Dyakonov theorem using the well-known Hardy-Littlewood theorem which says that an analytic function f on the unit disk satisfies (1.1) |f^{\prime}(z)|\leq C_{f}(1-|z|)^{\alpha-1},\quad z\in\mathbb{D}, if and only if it is \alpha -Hölder continuous, i.e. (1.2) |f(z)-f(w)|\leq\ C^{\prime}_{f}|z-w|^{\alpha},\quad z,w\in\mathbb{D}. Here, C_{f} and C^{\prime}_{f} are constants that depend only on f . Estimates of the type C^{\prime}_{f}\leq aC_{f} and C_{f}\leq bC^{\prime}_{f} , where a and b are absolute constants, are also known and important. After the Dyakonov result appeared, many authors considered whether a similar equivalence holds for other mapping classes, such as quasi-regular mappings or harmonic mappings on domains in \mathbb{R}^{n} . Our aim is to consider this question for mappings between metric spaces and, in particular, for analytic mappings on domains in a normed space of not necessarily finite dimension. The main results of this work are collected in the third section after we give some preliminaries and auxiliary results. Since a version of the Hardy-Littlewood theorem plays an important role in this paper, we shall present in the sequel some earlier extensions of it. The results obtained by Gehring and Martio and Lappalainen in a more general setting are those that served as motivation for the author’s generalization of the Hardy-Littlewood theorem for mappings between metric spaces [ 14 ] . Lappalainen [ 12 ] obtained an extension of the one part of the Hardy and Littlewood theorem for mappings on so called Lip φ -extension domains, where \varphi:[0,\infty)\to[0,\infty) is a majorant in the meaning: a continuous function, differentiable in (0,\infty) , with \varphi(0)=0 , \varphi(t)>0 for t\in(0,\infty) , an increasing function on (0,\infty) , and \varphi^{\prime} a decreasing function on (0,\infty) . In the definition of Lip φ -extension domains, an important role plays local Lipschitz calasses [ 12 ] . Roughly, a domain belongs to this class if every locally Lipschitz function is globally Lipschitz along with the norm control. A theorem says that a domain D with a non-empty boundary is a Lip φ -extension domain if and only if there exists a constant M such that for every x,y\in D one finds a rectifiable curve \gamma_{x,y}\subseteq D with ends at x and y such that (1.3) \int_{\gamma_{x,y}}\frac{\varphi(d(z,\partial D))}{d(z,\partial D)}|dz|\leq M\varphi(|x-y|), where d(z,\partial D) is the distance of z from \partial D . In the case of standard majorants \varphi(t)=\varphi_{\alpha}(t)=t^{\alpha} , \alpha\in(0,1) , it is said that a domain is a Lip α -extension domain, following the terminology of Gehring and Martio. A variant of the one part of the Hardy and Littlewood theorem obtained by Lappalainen [ 12 , Theorem 7.3] states that if an analytic function on a Lip φ -extension domain D\subseteq\mathbb{C} satisfies (1.4) |f^{\prime}(z)|\leq C_{f}\frac{\varphi(d(z,\partial D))}{d(z,\partial D)},\quad z\in D, then f satisfies the Lipschitz type condition (1.5) |f(z)-f(w)|\leq C^{\prime}_{f}\varphi(|z-w|),\quad z,w\in D, where C_{f} and C^{\prime}_{f} are constants. Gehring and Martio obtained a generalization of the same part of the Hardy and Littlewood theorem for mappings on uniform domains [ 6 ] , and somewhat later on Lip α -extension domains [ 7 ] . In this paper a variant of the Gehring and Martio result is proved for analytic mappings on uniform domains in a normed space. 2. Preliminaries A continuous mapping \gamma:[a,b]\to X , is called a curve in a metric space (X,d_{X}) that connects \gamma(a) and \gamma(b) . For simplicity, it is convenient to identify the curve \gamma with its image \gamma([a,b]) . In case \gamma being a 1-1 mapping, the curve is called an arc. The length of the curve \gamma is given by \ell(\gamma)=\sup\sum_{i=1}^{n}d_{X}(\gamma(t_{i-1}),\gamma(t_{i})), where the supremum is taken over all finite partitions a=t_{0}<t_{1}<\dots<t_{n}=b of [a,b] . A curve is rectifiable if its length is finite. The family of rectifiable curves that connect x\in X and y\in X is denoted by \Gamma({x,y}) . A metric space X is said to be rectifiable connected if \Gamma(x,y) is not empty for every x,y\in X . If \gamma:[a,b]\to X is a rectifiable curve in a metric space X , and [c,d]\subseteq[a,b] , then the (rectifiable) curve \gamma:[c,d]\to X is denoted by \gamma_{[c,d]} . In case \gamma is an arc, \gamma(c)=z , and \gamma(d)=w , we shall also use the notation \gamma[z,w] for this part of the curve. The integral of a continuous function f on X over a rectifiable curve \gamma\subseteq X is denoted by \int_{\gamma}f . It may be defined via generalized Riemann sums, as in [ 13 ] : For every \varepsilon>0 there exists \delta>0 such that \left|\int_{\gamma}f-\sum_{i=1}^{n}f(\gamma(s_{i}))\ell(\gamma|_{[t_{i-1},t_{i}]})\right|<\varepsilon for every subdivision a=t_{0}<t_{1}<\cdots<t_{n}=b of [a,b] with \max_{1\leq i\leq n}|t_{i}-t_{i-1}|<\delta , and every sequence s_{1},s_{2},\dots,s_{n} with t_{i-1}\leq s_{i}\leq t_{i} , 1\leq i\leq n . This definition is convenient for our consideration in the papers [ 13 ] and [ 14 ] . On the other hand, the definition may be given as in [ 10 , Chapter V] not only for continuous functions, using the fact that every rectifiable curve can be parameterized by the arc length. Let \gamma be a rectifiable curve and denote its length by \ell . For a Borel measurable function f in a metric space X one can define the integral of f over this curve by \int_{\gamma}f=\int_{0}^{\ell}f(\gamma(s))ds, where on the right side we have the standard Lebesgue integral. An everywhere positive and continuous function on a metric space is called a weight

Read the complete text