Limits at infinity for Hajłasz–Sobolev functions in metric spaces

Angha Agarwal, Antti V. Vähäkangas · 2025

Read Limits at infinity for Hajłasz–Sobolev functions in metric spaces by Angha Agarwal, Antti V. Vähäkangas (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-2506-05037.txt.

Introduction

Angha Agarwal and Antti Vähäkangas study function limits at infinity in a 2025 mathematics paper. The setting is abstract metric spaces with a doubling measure. Hajłasz-Sobolev functions generalize smooth functions to such spaces. Whether they settle to a limit far away was the question.

For uniformly perfect spaces, the answer is yes, almost everywhere at infinity. A well behaved representative of the function has a pointwise limit. The exceptions form a set measured by a variational capacity. Their framework sharpens older ones based on Hausdorff content. It also extends results for Newtonian and fractional Sobolev functions. Relative capacity gives a finer gauge of the bad set. The proof refines the approach rather than replacing it wholesale.

Catalog record

CREATORAngha Agarwal, Antti V. Vähäkangas
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE87 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-05037

Browse the full library · About the Tower of Babel

Full text

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

Limits at infinity for Hajłasz–Sobolev functions in metric spaces Thanks: The authors are grateful to Pekka Koskela for suggesting the topic of the paper and for fruitful discussions. A. A. was supported by the Research Council of Finland Centre of Excellence in Randomness and Structures (Project number 364210). Angha Agarwal, Antti V. Vähäkangas Abstract . We study limits at infinity for homogeneous Hajłasz–Sobolev functions defined on uniformly perfect metric spaces equipped with a doubling measure. We prove that a quasicontinuous representative of such a function has a pointwise limit at infinity outside an exceptional set, defined in terms of a variational relative capacity. Our framework refines earlier approaches that relied on Hausdorff content rather than relative capacity, and it extends previous results for homogeneous Newtonian and fractional Sobolev functions. 2020 Mathematics Subject Classification 46E36, 31C15, 31B15, 31B25 1. Introduction In this paper, we study limits at infinity for functions in the homogeneous Hajłasz–Sobolev space within the framework of metric measure spaces. Related problems have been studied for functions in the first-order Sobolev space in [ 6 , 7 , 22 , 23 , 29 , 31 ] . Similar questions have been addressed for functions of bounded variation in [ 25 ] , and for functions in fractional Sobolev spaces in [ 1 ] . We briefly review the results in [ 22 , 1 ] that are most relevant to our work. Kline, Koskela, and Nguyen [ 22 , Theorem 1.1] obtained limits at infinity for functions in the first-order homogeneous Newton–Sobolev space outside thin exceptional sets, which have sufficiently small Hausdorff contents in dyadic annuli at infinity. Their results rely on the assumption that the ambient metric measure space is complete, equipped with a d -regular measure, and supports a suitable Poincaré inequality. In a recent work, Agarwal, Koskela, and Mohanta [ 1 ] established the existence of limits at infinity for functions in homogeneous fractional Sobolev spaces, outside thin exceptional sets characterized by small Hausdorff content at infinity. In their setting, the metric space is {\mathbb{R}}^{n} equipped with the Euclidean metric and Lebesgue measure. Extending the approaches in [ 22 ] and [ 1 ] , we demonstrate the existence of limits at infinity for functions in the homogeneous Hajłasz–Sobolev space \dot{M}^{s,p}(X,d,\mu) , where (X,d,\mu) is a metric measure space, 0<s\leq 1 and 0<p<\infty . A priori, we do not assume that the underlying metric space X is connected. Instead, it is assumed to be uniformly perfect and equipped with a doubling measure \mu , providing the necessary geometric and measure-theoretic part of our framework described in Section 2 . The choice of the homogeneous Hajłasz–Sobolev space \dot{M}^{s,p} in our framework, defined in Section 4 , is general enough to encompass a wide range of Sobolev-type spaces. In particular, both the homogeneous Newton–Sobolev and fractional Sobolev spaces admit a bounded embedding into the homogeneous Hajłasz–Sobolev space considered here. For the former embedding to hold, one needs to assume a suitable Poincaré inequality as in [ 22 ] , forcing the space to be connected. We refer to Remark 8.6 for further discussion. The corresponding embedding for homogeneous fractional Sobolev spaces is developed in Section 9 . Throughout the paper, we use median values instead of integral averages to facilitate the analysis of functions in the homogeneous Hajłasz–Sobolev space \dot{M}^{s,p} . Such a function may fail to be integrable on balls for small values of s and p , hence the corresponding integral averages are not well-defined. The use of medians in this context is well established; we refer to [ 1 , 9 , 12 , 17 ] for related developments. See Sections 3 and 5 for a further discussion. A distinctive feature of our approach in studying limits at infinity is a novel use of variational relative capacity, as introduced in Definition 6.1 . In one of our main results, Theorem 7.9 , we establish conditions, under which a quasicontinuous representative of a function in the homogeneous Hajłasz–Sobolev space admits a limit of the form \lim_{\begin{subarray}{c}d(x,O)\to\infty\\ x\in X\setminus E(O)\end{subarray}}u(x)\,. Here O\in X is a fixed basepoint and E(O) is an exceptional set, outside of which the limit of the function u exists. To quantify the smallness of such an exceptional set at infinity, we utilize the variational relative capacity. These exceptional sets in our framework are called (s,p) -thin at infinity, and they are discussed in Section 7 . The novel use of relative capacity, instead of Hausdorff content as in [ 22 , 1 ] , provides a finer exceptional set, resulting in sharper and more precise analytical conclusions. We refer the reader to Section 8 and Theorem 8.4 for detailed comparisons. As an application, we extend and complement the results in [ 22 ] and [ 1 ] . For precise formulations, we refer to Remark 8.6 and Theorem 9.6 , respectively. 2. Metric spaces We introduce our standing assumptions on the metric space and the measure, which are maintained throughout the paper. For further details, we refer to [ 2 , 13 , 15 ] . Let (X,d,\mu) be a metric measure space with \operatorname{diam}(X)=\infty . Here \mu is a positive complete Borel measure on X such that 0<\mu(B(x,r))<\infty for all x\in X and r>0 , where B(x,r)=\{y\in X\,:\,d(y,x)<r\} is the open ball of radius r>0 and center x\in X . The measure \mu is assumed to be doubling, that is, there exists a constant c_{\mu}\geq 1 such that, whenever x\in X and r>0 , we have (1) \mu(B(x,2r))\leq c_{\mu}\,\mu(B(x,r)). We also assume that the measure \mu satisfies the reverse doubling condition, that is, there are constants 0<c_{R}<1 and \kappa>1 such that (2) \mu(B(x,r))\leq c_{R}\,\mu(B(x,\kappa r)) for all x\in X and all r>0 . The above assumptions are tacitly held throughout the paper. The space X is separable under these assumptions, see [ 2 , Proposition 1.6] . It follows from the reverse doubling condition ( 2 ) that \mu(X)=\infty . If X is connected, then ( 2 ) for any \kappa>1 follows from the doubling condition ( 1 ) and the unboundedness of X ; see for instance [ 2 , Lemma 3.7] . However, we do not assume in this paper that X is connected unless otherwise specified. Since \mu is doubling in our setting, the reverse doubling condition ( 2 ) is equivalent with the so-called uniform perfectness of X , we refer to [ 13 , Section 11] for background on this condition and examples. We also need the following iterated versions of the doubling and reverse doubling conditions. Iteration of the doubling condition ( 1 ) gives an exponent Q>0 and a constant c_{Q}>0 , both only depending on c_{\mu} , such that (3) \frac{\mu(B(y,r))}{\mu(B(x,r_{0}))}\geq c_{Q}\left(\frac{r}{r_{0}}\right)^{Q} whenever x\in X , y\in B(x,r_{0}) and 0<r<r_{0}<\infty . Condition ( 3 ) holds for Q=\log_{2}c_{\mu} , we refer to [ 2 , Lemma 3.3] for details. On the other hand, the doubling property and iteration of the reverse doubling condition ( 2 ) gives an exponent \sigma>0 and a constant c_{\sigma}>0 , both only depending on c_{R} , such that the quantitative reverse doubling condition (4) \frac{\mu(B(x,r))}{\mu(B(x,r_{0}))}\leq c_{\sigma}\left(\frac{r}{r_{0}}\right)^{\sigma} holds for all x\in X and 0<r<r_{0}<\infty . We refer to the proof

Read the complete text