Recognition: unknown
Weak convergence from projected laws on a positive-measure set of directions
Pith reviewed 2026-05-10 15:14 UTC · model grok-4.3
The pith
If the target one-dimensional projected laws are moment-determinate for almost every direction, then convergence of those projections on any positive-measure set of directions implies weak convergence of the measures on R^d.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
If the projected laws of the target measure are moment-determinate for surface-almost every direction, then weak convergence of a sequence of probability measures on R^d follows from convergence of the projected laws on any subset of directions that has positive surface measure on the sphere.
What carries the argument
The moment-determinacy of the one-dimensional projected laws for surface-almost every direction on the sphere, which upgrades convergence on a positive-measure subset to convergence on a dense enough collection to invoke the full Cramér-Wold device.
If this is right
- Weak convergence on R^d holds whenever the projected laws converge on any positive-measure subset of directions, once the moment-determinacy condition is met.
- The classical requirement of convergence in every direction can be replaced by convergence on a set of positive surface measure under the same assumption.
- For any probability distribution on the sphere that is absolutely continuous with respect to surface measure, convergence of the projected law along a randomly sampled direction implies weak convergence with probability one.
Where Pith is reading between the lines
- Numerical or statistical verification of weak convergence could be performed by checking only a random sample of directions rather than a dense grid.
- The same relaxation may apply to other notions of convergence that rely on one-dimensional marginals when uniqueness from moments holds almost everywhere.
Load-bearing premise
The one-dimensional projected laws of the target measure are moment-determinate for surface-almost every direction on the sphere.
What would settle it
An explicit sequence of measures on R^d whose one-dimensional projections converge on a positive-measure set of directions, yet the sequence fails to converge weakly, while the target projections fail to be moment-determinate on a positive-measure set of directions.
read the original abstract
The Cram\'er-Wold device characterises weak convergence of probability measures on $\mathbb{R}^d$ through convergence of all one-dimensional projected laws. We prove that, if the target projected laws are moment-determinate for surface-almost every direction, then weak convergence already follows from projected convergence on a positive-measure set of directions. This yields a simple probabilistic interpretation: if one samples a direction at random from any distribution on the sphere that is absolutely continuous with respect to surface measure, then, with probability one, convergence of the projected law along the sampled direction already forces global weak convergence under the same moment-determinacy assumption.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proves an extension of the Cramér-Wold theorem: if the one-dimensional projections of a target probability measure μ on R^d are moment-determinate for surface-almost every direction, then convergence of the projected laws on any positive surface-measure subset A of directions implies weak convergence of the sequence to μ. A probabilistic reading is given via almost-sure convergence under random sampling of directions absolutely continuous w.r.t. surface measure.
Significance. If the central claim holds, the result meaningfully weakens the directional requirement in Cramér-Wold while retaining a clean analytic proof via polynomial moment functions and the identity theorem. It is of interest for applications in high-dimensional probability and statistics where checking projections in all directions is infeasible. The manuscript correctly credits the role of moment-determinacy and provides an intuitive random-direction interpretation.
major comments (1)
- [§3] §3 (Proof of Theorem 1.1, the moment-agreement step): the argument that the difference of the k-th moment functions is a homogeneous polynomial vanishing on A and hence identically zero presupposes that every weak limit point λ has finite k-th moments in every direction. The hypothesis only guarantees finite moments (and agreement) on the positive-measure set A; no uniform integrability, tightness, or a-priori bound ensuring global moment existence for λ is stated. Without this, the polynomial representation and analytic continuation do not apply globally.
minor comments (2)
- [Abstract] Abstract: the phrase 'surface-almost every direction' is clear to specialists but would benefit from an explicit parenthetical reference to the normalized surface measure on the sphere.
- [§1] Notation: the symbol for the sphere and the surface measure could be introduced once in §1 and used consistently thereafter.
Simulated Author's Rebuttal
We thank the referee for the careful reading and for identifying a subtlety in the moment-agreement step of the proof of Theorem 1.1. The observation is well-taken and highlights the need for an explicit justification that weak limit points possess finite moments in all directions. We address the point below and will revise the manuscript to incorporate the necessary clarifications.
read point-by-point responses
-
Referee: [§3] §3 (Proof of Theorem 1.1, the moment-agreement step): the argument that the difference of the k-th moment functions is a homogeneous polynomial vanishing on A and hence identically zero presupposes that every weak limit point λ has finite k-th moments in every direction. The hypothesis only guarantees finite moments (and agreement) on the positive-measure set A; no uniform integrability, tightness, or a-priori bound ensuring global moment existence for λ is stated. Without this, the polynomial representation and analytic continuation do not apply globally.
Authors: We agree that the original write-up did not sufficiently detail why weak limit points λ have finite k-th moments in every direction. We will revise §3 by inserting the following preparatory arguments before the moment-agreement step. First, the sequence {μ_n} is tight: if not, some subsequence would send positive mass to infinity along a direction v; the projected laws would then fail to be tight for all θ with ⟨v, θ⟩ ≠ 0, i.e., for surface-almost every θ. This contradicts the assumed weak convergence (hence tightness) of the projections on the positive-measure set A. Tightness and Prokhorov’s theorem yield weak limit points λ. Second, weak convergence in R^d implies, by the continuous-mapping theorem, that the one-dimensional projections converge weakly in every direction. Consequently, on A the projections of λ coincide with those of μ and therefore possess finite moments. Third, the moment function f_k(θ) := ∫ |⟨x, θ⟩|^k dλ(x) is lower semi-continuous on the sphere. Moreover, if f_k(θ_0) = ∞ for any θ_0, the contribution of the tails along θ_0 forces f_k(θ) = ∞ for almost every θ (the equatorial set {θ : ⟨θ_0, θ⟩ = 0} has measure zero). Hence the set where f_k < ∞ is either null or conull. Since it contains the positive-measure set A, f_k is finite almost everywhere. Lower semi-continuity together with the directional-tail argument then implies finiteness everywhere. With moments finite in all directions, the difference of the k-th moment functions is a well-defined homogeneous polynomial that vanishes on A and is therefore identically zero by the identity theorem. The remainder of the proof proceeds unchanged. revision: yes
Circularity Check
No circularity; direct extension of external Cramér-Wold theorem
full rationale
The paper establishes that projected convergence on a positive-measure set of directions implies weak convergence when the target measures are moment-determinate for surface-almost every direction. This is positioned as an extension of the standard Cramér-Wold theorem using analytic properties of moment functions (homogeneous polynomials) and the identity theorem. No self-citations are load-bearing for the central claim, no parameters are fitted from data and then relabeled as predictions, and no definitions or ansatzes are smuggled in via prior work by the same authors. The derivation chain relies on external classical results and does not reduce to its own inputs by construction.
Axiom & Free-Parameter Ledger
axioms (2)
- standard math Cramér-Wold theorem: weak convergence holds iff all one-dimensional projections converge
- standard math Moment-determinate distributions are uniquely determined by their moments
Reference graph
Works this paper leans on
-
[1]
Cramér, H
H. Cramér, H. Wold, Some theorems on distribution functions,J. Lond. Math. Soc.11(4) (1936) 290–294
1936
-
[2]
Cuesta-Albertos, R
J.A. Cuesta-Albertos, R. Fraiman, T. Ransford, A sharp form of the Cramér–Wold theorem,J. Theoret. Probab.20(2) (2007) 201–209
2007
-
[3]
Cuesta-Albertos, E
J.A. Cuesta-Albertos, E. del Barrio, R. Fraiman, C. Matrán, The random projection method in goodness of fit for functional data,Comput. Statist. Data Anal.51(10) (2007) 4814–4831
2007
-
[4]
Petersen, On the relation between the multidimensional moment problem and the one-dimensional moment problem,Math
L.C. Petersen, On the relation between the multidimensional moment problem and the one-dimensional moment problem,Math. Scand.51(2) (1982) 361–366
1982
-
[5]
Shohat, J.D
J.A. Shohat, J.D. Tamarkin,The Problem of Moments, Mathematical Surveys and Monographs, vol. 1, American Mathematical Society, Providence, RI, 1943. 4
1943
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.