arxiv: 2605.03130 · v1 · submitted 2026-05-04 · 🧮 math.FA

Recognition: unknown

Image transformations, Markov operators, and sample median

S. V. Butler

Pith reviewed 2026-05-08 02:28 UTC · model grok-4.3

classification 🧮 math.FA

keywords Markov operatorsdeficient topological measuresimage transformationsKantorovich-Rubinstein metricsample medianiterated function systemsRadon measurescovering dimension

0 comments

The pith

Markov operators from contracting image transformations admit unique invariant deficient topological measures, and signed topological measures on spaces of dimension at most one are Radon measures.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper generalizes iterated function systems by defining Markov operators on spaces of deficient topological measures as infinite convex combinations of adjoints of image transformations. When these transformations contract distances in the Kantorovich-Rubinstein metric, the operator has a unique invariant deficient topological measure. It introduces the generalized distribution of the sample median for continuous proper maps and shows that this distribution together with its inverse is equivariant under solid variables such as rotations, translations, projections, and monotone maps on Euclidean space. The work further proves that any signed topological measure on a locally compact space of covering dimension at most one is necessarily a signed Radon measure. These statements recover the classical attractor result for iterated function systems as a special case when the space is compact and the operator is restricted to ordinary measures.

Core claim

If (d-) image transformations are contractions with respect to the Kantorovich-Rubinstein metric, then the associated Markov operator on deficient topological measures has a unique invariant measure. The generalized distribution of the sample median and the inverse on the sample median are equivariant under solid variables. A signed topological measure on a locally compact space with covering dimension at most one is a signed Radon measure.

What carries the argument

Markov operator defined as an infinite convex linear combination of adjoints of (d-) image transformations, with the Kantorovich-Rubinstein metric supplying the contraction condition that guarantees uniqueness of the invariant deficient topological measure.

Load-bearing premise

The spaces are locally compact, the maps are continuous and proper, the image transformations satisfy contraction conditions in the Kantorovich-Rubinstein metric, and the covering dimension is at most one for the Radon-measure identification.

What would settle it

An explicit collection of continuous proper maps that contract in the Kantorovich-Rubinstein metric yet induce a Markov operator without a unique invariant deficient topological measure, or a signed topological measure on the circle that is not a signed Radon measure.

read the original abstract

(I.) We consider generalizations of an iterated function system and the associated Markov operators. A Markov operator, defined on the space of (deficient) topological measures on a locally compact space, is an infinite convex linear combination of adjoints of (d-) image transformations. Restricted to measures, this Markov-Feller operator has a nonlinear dual operator given by an infinite convex linear combination of (conic) quasi-homomorphisms. If (d-) image transformations are contractions with respect to the Kantorovich-Rubinstein metric, a Markov operator has the unique invariant (deficient) topological measure. Taking a compact space, finitely many inverses of contractions as image transformations, and restricting the Markov operator to measures gives the classical result from the theory of fractals. There are various relations between Markov operator and the iterated function system where adjoints of (d-) image transformations are contractions on the compact metric space of $\{0,1\}$-valued (deficient) topological measures. For instance, the invariant (deficient) topological measure is the composition of the fixed point of the IFS and the basic (d-) image transformation. (II.) We define a generalized distribution of the sample median (g.d.s.m.) for continuous proper maps using an image transformation. We show that the g.d.s.m. and the inverse on the sample median are equivariant under solid variables, a large collection of transformations. On $\mathbb{R}^n$ such transformations include rotations, translations, symmetries, stretching, projections, monotone maps, etc. (III.) We show that a (signed) topological measure on a locally compact space with the covering dimension $\dim X \le 1$ is a (signed) Radon measure.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper extends Markov operators to deficient topological measures via image transformations, adds an equivariant generalized sample median, and identifies topological measures with Radon measures when dimension is at most 1, but the abstract leaves the actual derivations unshown.

read the letter

The main things here are the move from standard IFS to Markov operators on deficient topological measures, the definition of a generalized sample-median distribution that stays equivariant under solid transformations like rotations and projections, and the claim that signed topological measures on spaces with covering dimension at most 1 are just signed Radon measures. The first part recovers the usual fractal fixed-point result as a special case when you restrict to ordinary measures and finite contractions. The equivariance section lists a useful collection of maps that preserve the construction, which could be handy for someone working with medians on transformed data. The dimension bound is presented as a clean identification theorem. These pieces sit inside the usual Kantorovich-Rubinstein contraction setup and the standard axioms for topological measures, so the framework is consistent with existing tools rather than a sharp break. The relations between the adjoint operators and the IFS fixed point are laid out explicitly, which helps tie the pieces together. The soft spots are the absence of any proof outlines or counter-example checks in the summary; the uniqueness of the invariant deficient measure and the nonlinear dual operator rest on the contraction assumption, but without the steps it is hard to judge whether extra conditions slipped in. The dimension-1 result looks plausible from the literature on topological measures, yet it would need a direct check to confirm it is not just a restatement. This is for readers already comfortable with IFS theory and generalized measures who want to see how sample medians fit into the same operator picture. It is narrow enough that most people outside that subfield will not need it, but the constructions are concrete enough that a specialist could test them. The work deserves a serious referee to look at the full arguments and the citation record for the measure identification.

Referee Report

2 major / 2 minor

Summary. The manuscript develops three main results in functional analysis: (I) generalizations of iterated function systems via Markov operators on spaces of (deficient) topological measures on locally compact spaces, where the operator is an infinite convex combination of adjoints of (d-) image transformations; under the assumption that these transformations are contractions w.r.t. the Kantorovich-Rubinstein metric, the Markov operator admits a unique invariant (deficient) topological measure, recovering the classical IFS result on compact spaces when restricted to measures; (II) the definition of a generalized distribution of the sample median (g.d.s.m.) via image transformations for continuous proper maps, together with equivariance of the g.d.s.m. and the inverse on the sample median under solid variables (including rotations, translations, symmetries, etc., on R^n); (III) the identification of (signed) topological measures with (signed) Radon measures on locally compact spaces of covering dimension at most 1.

Significance. If the derivations hold, the work provides a natural extension of IFS theory and Markov-Feller operators to deficient topological measures using the Kantorovich-Rubinstein metric and adjoints, without introducing free parameters or circular constructions. The equivariance results for the sample-median distribution offer a new tool with potential statistical and imaging applications, while the dimension-1 characterization clarifies the relationship between topological and Radon measures under standard local-compactness and continuity assumptions. The paper explicitly invokes the usual minimal hypotheses (local compactness, proper continuous maps, contractions, dim X ≤ 1) and builds on established axioms.

major comments (2)

[Abstract (I)] Abstract (I): the uniqueness of the invariant deficient topological measure is asserted when image transformations are KR-contractions, but the abstract supplies no proof sketch, error estimates, or explicit invocation of the Banach fixed-point theorem on the space of deficient topological measures; the full manuscript must supply this derivation (or a reference to a prior result) as it is load-bearing for the central claim.
[Abstract (II)] Abstract (II): the g.d.s.m. is defined via an image transformation and claimed to be equivariant under solid variables, yet no explicit verification or counter-example check is indicated in the abstract; the manuscript should include the precise definition of the g.d.s.m. and the step-by-step equivariance argument for at least one non-trivial solid variable (e.g., a projection or monotone map) to substantiate the claim.

minor comments (2)

[Introduction] The term 'deficient topological measure' is used throughout but its precise axiomatic definition (and how it differs from a standard topological measure) should be stated explicitly in the introduction or a preliminary section.
[Notation] Notation for 'd- image transformations' and 'solid variables' should be defined at first use and consistently employed; a short table or list of examples for solid variables on R^n would improve readability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed reading and the constructive suggestions. We address each major comment below. The full manuscript already contains the required derivations; we will expand the abstract and add an explicit verification example to make the claims more self-contained.

read point-by-point responses

Referee: [Abstract (I)] Abstract (I): the uniqueness of the invariant deficient topological measure is asserted when image transformations are KR-contractions, but the abstract supplies no proof sketch, error estimates, or explicit invocation of the Banach fixed-point theorem on the space of deficient topological measures; the full manuscript must supply this derivation (or a reference to a prior result) as it is load-bearing for the central claim.

Authors: The full manuscript (Section 3) derives uniqueness by showing that the Markov operator is a contraction mapping on the complete metric space of deficient topological measures equipped with the Kantorovich-Rubinstein metric; the contraction constant equals the supremum of the Lipschitz constants of the image transformations. The Banach fixed-point theorem is invoked directly to obtain the unique fixed point. We will insert a one-sentence proof sketch and the explicit reference to the Banach theorem into the abstract. revision: yes
Referee: [Abstract (II)] Abstract (II): the g.d.s.m. is defined via an image transformation and claimed to be equivariant under solid variables, yet no explicit verification or counter-example check is indicated in the abstract; the manuscript should include the precise definition of the g.d.s.m. and the step-by-step equivariance argument for at least one non-trivial solid variable (e.g., a projection or monotone map) to substantiate the claim.

Authors: Section 4 defines the g.d.s.m. as the push-forward of the empirical measure under the image transformation associated to the continuous proper map. Equivariance follows because every solid variable commutes with the image transformation (by the definition of solid variables). We will add the precise definition to the abstract and include a short step-by-step verification for the case of an orthogonal projection onto a coordinate subspace. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivations are self-contained extensions of standard results

full rationale

The paper applies the Banach fixed-point theorem to Markov operators constructed as convex combinations of adjoints of contractions with respect to the Kantorovich-Rubinstein metric on the space of deficient topological measures. This yields uniqueness of the invariant measure as a direct consequence of the contraction hypothesis and local compactness, without any parameter fitting or self-referential definitions. The equivariance of the generalized distribution of the sample median under solid variables follows from the definition of image transformations and proper continuous maps. The identification of signed topological measures with Radon measures when dim X ≤ 1 is stated as a theorem under the given hypotheses and recovers the classical IFS result on compact spaces as a special case rather than presupposing it. No load-bearing self-citations, ansatzes smuggled via prior work, or renamings of known patterns appear in the derivation chain.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 2 invented entities

Assessment limited to abstract; standard background axioms of locally compact spaces, continuous proper maps, and the Kantorovich-Rubinstein metric are presupposed but not enumerated. No free parameters or invented entities with independent evidence are visible.

axioms (2)

domain assumption Locally compact spaces admit well-behaved deficient topological measures and continuous proper maps
Invoked in the definitions of image transformations and Markov operators throughout parts (I) and (II).
domain assumption Covering dimension ≤1 implies topological measures coincide with Radon measures
Central to the statement in part (III); no proof details given.

invented entities (2)

deficient topological measure no independent evidence
purpose: Generalize ordinary measures to support the Markov-operator framework on non-compact spaces
Introduced as the domain for the generalized Markov operators in part (I).
generalized distribution of the sample median (g.d.s.m.) no independent evidence
purpose: Extend the classical sample median to continuous proper maps via image transformations
New object defined and shown equivariant in part (II).

pith-pipeline@v0.9.0 · 5605 in / 1715 out tokens · 58408 ms · 2026-05-08T02:28:32.651556+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

46 extracted references · 21 canonical work pages

[1]

F.: Quasi-states and quasi-measures

Aarnes, J. F.: Quasi-states and quasi-measures. Adv. Math. 86 no. 1, 41–67 (1991)

1991
[2]

F.: Pure quasi-states and extremal quasi-measures

Aarnes, J. F.: Pure quasi-states and extremal quasi-measures. Math. Ann. 295, 575–588 (1993) 29

1993
[3]

F.: Image transformstions, attractors, and invariant non- subadditive measures

Aarnes, J. F.: Image transformstions, attractors, and invariant non- subadditive measures. Preprint Mathematics Trondheim 2, 1–28 (1994)

1994
[4]

F.: Construction of non-subadditive measures and discretization of Borel measures

Aarnes, J. F.: Construction of non-subadditive measures and discretization of Borel measures. Fundam. Math. 147, 213–237 (1995)

1995
[5]

F., Grubb, D

Aarnes, J. F., Grubb, D. J.: Quasi-measures and image transforma- tion on completely regular spaces. Topol. Appl. 135 (1-3), 33–46 (2004). https://doi.org/10.1016/S0166-8641(03)00133-0

work page doi:10.1016/s0166-8641(03)00133-0 2004
[6]

F., Johansen, O., Rustad, A

Aarnes, J. F., Johansen, O., Rustad, A. B.: Topological measures, image transformations and self-similar sets. Acta Math. Hung. 109(1-2), 65–97 (2005). DOI:10.1007/s10474-005-0235-6

work page doi:10.1007/s10474-005-0235-6 2005
[7]

Aarnes, J., Rustad, A.: Probability and quasi-measures–a new interpretation. Math. Scand. 85 (2), 278–284 (1999). https://doi.org/10.7146/math.scand.a- 18277

work page doi:10.7146/math.scand.a- 1999
[8]

Random Comput

˚Akerlund-Bistr¨ om, C.: A generalization of the Hutchinson distance and applications. Random Comput. Dyn. 5(2&3), 159–176 (1997)

1997
[9]

Andreou, T.: Density of Lipschitz functions. Proceedings of The Conference of Applied Differential Geometry-General Relativity and The Workshop on GlobalAnalysis,DifferentialGeometryandLieAlgebras2001,BalkanSociety of Geometers, Geometry Balkan Press, 1–4 (2004)

2004
[10]

F.: Fractals Everywhere

Barnsley, M. F.: Fractals Everywhere. Second edition. Academic Press, Boston (1993)

1993
[11]

Bogachev, V. I. : Measure Theory. vol. 2: Regular and Chaotic Dynamics, Izhevsk (2003). English transl.: Springer-Verlag, Berlin (2007)

2003
[12]

Bogachev,V.I.,Kolesnikov,A.V.:TheMonge-Kantorovichproblem:achieve- ments, connections, and perspectives. Russ. Math. Surv., 67 (5) ( 2012). DOI 10.1070/RM2012v067n05ABEH004808

work page doi:10.1070/rm2012v067n05abeh004808 2012
[13]

Butler, S.: q-Functions and extreme topological measures. J. Math. Anal. Appl. 307, 465–479 (2005). https://doi.org/10.1016/j.jmaa.2005.01.013

work page doi:10.1016/j.jmaa.2005.01.013 2005
[14]

Mathematics

Butler, S. V.: Ways of obtaining topological measures on locally compact spaces. Bull. Irkutsk State Univ., Series "Mathematics" (Izvestiya Irkutskogo Gosudarstvennogo Universiteta, Seriya "Matematika"). 25, 33–45 (2018). https://doi.org/10.26516/1997-7670.2018.25.33

work page doi:10.26516/1997-7670.2018.25.33 2018
[15]

V.: Semisolid sets and topological measures

Butler, S. V.: Semisolid sets and topological measures. Topol. Appl. 310, 1–34 (2022). https://doi.org/10.1016/j.topol.2022.108036 30

work page doi:10.1016/j.topol.2022.108036 2022
[16]

V.: Non-linear functionals, deficient topological measures, and rep- resentation theorems on locally compact spaces

Butler, S. V.: Non-linear functionals, deficient topological measures, and rep- resentation theorems on locally compact spaces. Banach J. Math. Anal. 14 (3), 674–706 (2020). https://doi.org/10.1007/s43037-019-00034-0

work page doi:10.1007/s43037-019-00034-0 2020
[17]

V.: Quasi-linear functionals on locally compact spaces

Butler, S. V.: Quasi-linear functionals on locally compact spaces. Confluentes Mathematici. 13 (1), 3-34 (2021). DOI:10.5802/cml.69

work page doi:10.5802/cml.69 2021
[18]

V.: Deficient topological measures on locally compact spaces

Butler, S. V.: Deficient topological measures on locally compact spaces. Math. Nachr. 294 (6), 1115–1133 (2021). https://doi.org/10.1002/mana.201800574

work page doi:10.1002/mana.201800574 2021
[19]

Butler,S.V.:Signedtopologicalmeasuresonlocallycompactspaces.Analysis Math., 45 (4), 757–773 (2019)

2019
[20]

Butler, S. V. : Decompositions of signed deficient topological measures. Letters in Math Sci. 1, 61–77 (2019)

2019
[21]

V.: Integration with respect to deficient topological measures on locally compact spaces

Butler, S. V.: Integration with respect to deficient topological measures on locally compact spaces. Math. Slovaca 70 (5), 1113–1134 (2020). https://doi.org/10.1515/ms-2017-0418

work page doi:10.1515/ms-2017-0418 2020
[22]

V.: Weak Convergence of Topological Measures

Butler, S. V.: Weak Convergence of Topological Measures. J. Theor. Probab. 35, 1614–1639 (2022). https://doi.org/10.1007/s10959-021-01095-4

work page doi:10.1007/s10959-021-01095-4 2022
[23]

arXiv:2501.10635, 1–40 (2025)

Butler, S.: Quasi-linear maps and image transformations. arXiv:2501.10635, 1–40 (2025)

work page arXiv 2025
[24]

J.: Signed Quasi-measures and Dimension Theory

Grubb, D. J.: Signed Quasi-measures and Dimension Theory. Proc. Am. Math. Soc. 128 (4), 1105-1108 (2000). DOI:10.2307/119785

work page doi:10.2307/119785 2000
[25]

Dickstein, A., Zapolsky, F.: Approximation of quasi-states on manifolds. J. of Appl. and Computational Top. 3, 221–248 (2019)

2019
[26]

John Wiley&sons, Chichester (1997)

Falconer, K.: Techniques in fractal geometry. John Wiley&sons, Chichester (1997)

1997
[27]

F.: What is the Center of an Area, or the Center of a Population? Publications of the American Statistical Association vol

Hayford, J. F.: What is the Center of an Area, or the Center of a Population? Publications of the American Statistical Association vol. 8 (58), 1–12 (1902)

1902
[28]

Indiana Univ

Hutchinson, J: Fractals and self similarity. Indiana Univ. Math. J. 30 (5), 713–747 (1981)

1981
[29]

V.: On the translocation on masses

Kantorovich, L. V.: On the translocation on masses. Dokl. akad. Nauk SSSR, vol. 37. (7–8) 227–229 (1942) (in Russian). English transl.: C. R. (Doklady) Acad. Sci. URSS, vol. 37, 199–201 (1942)

1942
[30]

S.: Completeness of the space of separable measures in the Kantorovich-Rubinshtein metric

Kravchenko, A. S.: Completeness of the space of separable measures in the Kantorovich-Rubinshtein metric. Sib. Math. J. 47 (1), 68–76 (2006) 31

2006
[31]

B.: Construction and Properties of quasi- linear functionals

Johansen,Ø., Rustad, A. B.: Construction and Properties of quasi- linear functionals. Trans. Am. Math. Soc. 358 (6), 2735–2758 (2006). DOI:10.2307/3845505

work page doi:10.2307/3845505 2006
[32]

B.: The homology of spaces of simple topological measures

Johansen,Ø., Rustad, A. B.: The homology of spaces of simple topological measures. Fundam. Math. 177(1), 19–43 (2003). DOI:10.4064/fm177-1-2

work page doi:10.4064/fm177-1-2 2003
[33]

Rustad, A.: Constructing topological measures and Image transformations by Boolean functions

Johansen,Ø., A. Rustad, A.: Constructing topological measures and Image transformations by Boolean functions. Preprint, 1–19, (2003)

2003
[34]

Houston J

Laberge, T.: Supports of quasi-measures. Houston J. Math. 24 (2), 301–312 (1998)

1998
[35]

Lasota, A., Myjak, J., Szarek, T.: Markov operators with a unique invariant measure. J. Math. Anal. Appl. 276, 343–356 (2002). https://doi.org/10.1016/S0022-247X(02)00457-2

work page doi:10.1016/s0022-247x(02)00457-2 2002
[36]

Dissertation, NTNU, Nov

Pedersen,Ø: Topological measures and Image Transformations. Dissertation, NTNU, Nov. 2002

2002
[37]

CRM Monograph series, vol

Polterovich, L., Rosen, D.: Function theory on symplectic manifolds. CRM Monograph series, vol. 34. AMS, Providence RI (2014)

2014
[38]

B.: The multidimensional median as a quasi-measure

Rustad, A. B.: The multidimensional median as a quasi-measure. Preprint NTNU, Mathematics. no. 5, 1–27 (1998)

1998
[39]

B.: Unbounded quasi-integrals

Rustad, A. B.: Unbounded quasi-integrals. Proc. Am. Math. Soc. 129 (1), 165–172 (2000). DOI:10.2307/2669042

work page doi:10.2307/2669042 2000
[40]

Numerical experiments in revisited brittle fracture

Rustad, A. B.: Quasi-measures with image transformations as generalized variables. J. Math. Anal. Appl. 271, 16–30 (2002). DOI:10.1016/S0022- 247X(02)00076-8

work page doi:10.1016/s0022- 2002
[41]

B.: The median in multidimensional spaces

Rustad, A. B.: The median in multidimensional spaces. Adv. Appl. Math. 33 (2), 366–396 (2004). https://doi.org/10.1016/j.aam.2003.10.001

work page doi:10.1016/j.aam.2003.10.001 2004
[42]

G.: A Survey of Multidimensional Medians

Small, C. G.: A Survey of Multidimensional Medians. International Statistics Review/Revue Internationale de Statistique, 58 (3), 263–277 (1990)

1990
[43]

G.: A Signed quasi-measure decomposition

Svistula, M. G.: A Signed quasi-measure decomposition. Vestn. Samar. Gos. Univ. Estestvennonauchn. 62 (3), 192–207 (2008) (in Russian)

2008
[44]

G.: Deficient topological measures and function- als generated by them

Svistula, M. G.: Deficient topological measures and function- als generated by them. Sb.: Math. 204 (5), 726–761 (2013). DOI:10.1070/SM2013v204n05ABEH004318 32

work page doi:10.1070/sm2013v204n05abeh004318 2013
[45]

N.: Markovskie protsessy na schetnom proizvedenii pros- transtv, opisyvaiushchie bolshie sistemy avtomatov

Vasershtein, L. N.: Markovskie protsessy na schetnom proizvedenii pros- transtv, opisyvaiushchie bolshie sistemy avtomatov. Probl. peredachi inform., 5(3),64–72(1969).Englishtransl.:L.N.Vasershtein,L.N.:Markovprocesses over denumerable products of spaces describing large system of automata. Problems Inform. Transmission, 5 (3) 47–52 (1969)

1969
[46]

F.: Quasi-measures and dimension theory

Wheeler, R. F.: Quasi-measures and dimension theory. Topology Appl. 66, 75-92 (1995) 33

1995