Generalized upper and lower Legendre conjugates for weight functions

Gerhard Schindl · 2025

Read Generalized upper and lower Legendre conjugates for weight functions by Gerhard Schindl (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-2505-07497.txt.

Introduction

Gerhard Schindl introduces new function transformations in a 2025 mathematics paper. The transforms extend the classic Legendre conjugate, a tool that swaps a function for a slope-based partner. Schindl defines upper and lower versions that work for pairs of functions with basic growth behavior.

He then studies how the transforms change growth indices that were defined in recent work on weight functions. A key case uses weight functions built from an underlying weight sequence. There, the transforms amount to multiplying or dividing the sequences term by term. Schindl argues this clean match shows the definitions are natural. He also points to uses in weighted function spaces in several branches of analysis.

Catalog record

CREATORGerhard Schindl
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE132 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-07497

Browse the full library · About the Tower of Babel

Full text

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

Generalized upper and lower Legendre conjugates for weight functions Thanks: This research was funded in whole or in part by the Austrian Science Fund (FWF) 10.55776/PAT9445424 Gerhard Schindl Abstract . We introduce and study new transformations between two functions satisfying some basic growth properties and generalize the known lower and upper Legendre conjugate (or envelope). We also investigate how these transformations modify recently defined growth indices for weight functions. A special but important and useful situation, to which the knowledge is then applied, is when considering associated weight functions which are expressed in terms of an underlying weight sequence. In this case these transformations precisely correspond to the point-wise product resp. point-wise division of the given sequences. Therefore, the new approach studied in this work illustrates the genuineness and importance and suggests applications for weighted spaces in different directions. 2020 Mathematics Subject Classification 26A12, 26A48, 26A51 1. Introduction Weighted spaces appear in different contexts and fields in Mathematics, like in Functional Analysis and the theory of partial differential operators but also in applications in various directions. We mention ultradifferentiable and ultraholomorphic classes, see [ Komatsu73 ] , [ BraunMeiseTaylor90 ] , [ BonetMeiseMelikhov07 ] , [ compositionpaper ] , and [ Thilliezdivision ] , [ optimalflat23 ] ; weighted spaces of sequences of complex numbers or weighted spaces of formal power series, see [ petzsche ] and [ Borelmapalgebraity ] ; generalized Gelfand-Shilov classes, see [ nuclearglobal1 ] , [ nuclearglobal2 ] , [ GelfandShilovincl24 ] ; weighted spaces of entire functions, see [ Bonet2022survey ] , [ weightedentireinclusion1 ] and [ weightedentireinclusion2 ] ; and Orlicz spaces, see [ orliczpaper ] . In each instance we also refer to the list of references in these works. Classically, one can find two approaches: either using a weight sequence \mathbf{M}=(M_{p})_{p\in\mathbb{N}} or a weight function \omega:[0,+\infty)\rightarrow[0,+\infty) . More recently, even weight matrices or weight function systems have been considered; see e.g. [ compositionpaper ] , [ GelfandShilovincl24 ] . In all these settings growth and regularity assumptions on the weights are required and unavoidable. In general, the weight sequence and weight function setting are mutually distinct which follows in the ultradifferentiable case by [ BonetMeiseMelikhov07 ] , see also [ compositionpaper , Sect. 5] , and in the Gelfand-Shilov setting we refer to [ GelfandShilovincl , Sect. 5] . Note that in the mentioned settings similar growth conditions appear; this has motivated the abstract study in [ index ] where especially weight functions have been investigated under very general growth assumptions. Given a weight sequence \mathbf{M}=(M_{p})_{p\in\mathbb{N}}\in\mathbb{R}_{>0}^{\mathbb{N}} satisfying basic growth properties one can consider the corresponding associated function \omega_{\mathbf{M}} ; see [ mandelbrojtbook , Chapitre I] , [ Komatsu73 , Sect. 3] , the more recent work [ regularnew ] and Section 2.5 below for more details. If \mathbf{M} is log-convex (see ( 2.1 )), then there exists an intimate relation between \mathbf{M} and \omega_{\mathbf{M}} . Note that \omega_{\mathbf{M}} appears frequently in both the weight sequence and weight function setting and in the latter case it is also serving as a (counter-)example and is becoming relevant for comparison results between these settings; see [ BonetMeiseMelikhov07 ] . On the other hand, for abstractly given weight functions, again satisfying basic growth restrictions, one can assign an associated weight sequence, see [ weightedentireinclusion1 , Sect. 2.7] and [ orliczpaper , Sect. 5] , or even a weight matrix, see [ compositionpaper , Sect. 5] and [ dissertation ] . From an abstract point of view but also for concrete applications and when working with weighted structures one is interested in the following questions: (I) Let sequences \mathbf{M} , \mathbf{N} be given and write \mathbf{M}\cdot\mathbf{N}:=(M_{p}N_{p})_{p\in\mathbb{N}} and \frac{\mathbf{M}}{\mathbf{N}}:=\left(\frac{M_{p}}{N_{p}}\right)_{p\in\mathbb{N}} for the point-wise product resp. point-wise quotient sequence. How is then \omega_{\mathbf{M}\cdot\mathbf{N}} , \omega_{\frac{\mathbf{M}}{\mathbf{N}}} related to \omega_{\mathbf{M}} , \omega_{\mathbf{N}} ? (II) Study for more general resp. for a wider class of weight functions these analogous relations and transformations; more precisely one may ask: Given functions \sigma,\tau:[0,+\infty)\rightarrow[0,+\infty) , which function is then corresponding to the operation yielding \omega_{\mathbf{M}\cdot\mathbf{N}} resp. \omega_{\frac{\mathbf{M}}{\mathbf{N}}} ? The content of this work is to answer and study (I) and (II) in detail: It turns out that taking the point-wise product is corresponding to (1.1) \sigma\check{\star}\tau(t):=\inf_{s>0}\{\sigma(s)+\tau(t/s)\},\;\;\;t\in[0,+\infty), whereas the point-wise quotient is corresponding to (1.2) \sigma\widehat{\star}\tau(t):=\sup_{s\geq 0}\{\sigma(s)-\tau(s/t)\},\;\;\;t\in(0,+\infty). The choice \tau=\operatorname{id} in ( 1.1 ) gives, up to inversion of the variable, the known lower Legendre conjugate (or envelope) of \sigma and ( 1.2 ) for \tau=\operatorname{id} gives the upper Legendre conjugate (or envelope) of \sigma ; see e.g. [ index , Sect. 2.5] and the references therein and confirm formulas ( 3.3 ) and ( 4.9 ). It is natural to study (II) for weight functions in the sense of [ index ] , see Definition 2.2 , and this notion encompasses in particular weight functions in the sense of Braun-Meise-Taylor introduced in [ BraunMeiseTaylor90 ] . In [ index ] for weight functions also new growth indices \gamma(\cdot) and \overline{\gamma}(\cdot) have been introduced, see Section 2.4 , and hence it is also natural to ask how these indices are modified under the actions \check{\star} and \widehat{\star} . Indeed, the starting point of our investigations and interest has been the technical statements shown in [ sectorialextensions , Sect. 3] . There we have revisited the lower/upper Legendre conjugate and established a connection between these conjugates and the associated functions \omega_{\frac{\mathbf{M}}{\mathbf{G}^{s}}} resp. \omega_{\mathbf{M}\cdot\mathbf{G}^{1}} , \mathbf{G}^{s}:=(p!^{s})_{p\in\mathbb{N}} , s>0 , see [ sectorialextensions , Lemmas 3.1, 3.4 & 3.5] . Note that \omega_{\mathbf{G}^{1}} corresponds (up to equivalence) to the weight \operatorname{id}:t\mapsto t and taking the power of a sequence is yielding a power substitution for the associated function (recall [ sectorialextensions , (2.7) ] ). Moreover, in the ultraholomorphic setting (for proving extension results) the point-wise product resp. point-wise quotient of a sequence \mathbf{M} and the Gevrey sequence \mathbf{G}^{s} appears naturally and is intimately related to the (formal) s -Laplace resp. s -Borel transform; we refer to [ surjectivityregularsequence , Thm. 3.5] and the citations and explanations stated just before this result. Then we have found [ Boman98 , Lemma 4] where an explicit relation between \omega_{\mathbf{M}\cdot\mathbf{N}} and \omega_{\mathbf{M}} , \omega_{\mathbf{N}} has been established in terms of ( 1.1 ). This operation is called “infimal convolution” in [ Boman98 ] and in view of this \widehat{\star} can be called “supremal anti-convolution” but these terminologies will not be used in this work. To the best of our knowledge the analogous result for \omega_{\frac{\mathbf{M}}{\mathbf{N}}} , when using ( 1.2 ), has not been shown so far and it seems that formally the proof given in [ Boman98 , Lemma 4] is not correct since the assumptions for \mathbf{M} , \mathbf{N} there do not ensure necessarily the required property that the corresponding sequences of quotients tend to infinity. In order to ensure that \sigma\widehat{\star}\tau and \omega_{\frac{\mathbf{M}}{\mathbf{N}}} are well-defined some more technical investigations have to be taken into account and the weight functions resp. sequences have to satisfy a certain growth relation; see Sections 4 and 5.4 for details. Roughly speaking this is due to the fact that division destroys regularity; i.e. even if both \mathbf{M} and \mathbf{N} are very regular then the point-wise quotient \frac{\mathbf{M}}{\mathbf{N}} will be not in general. Moreover, in the weight

Read the complete text