Pith. sign in

REVIEW 3 major objections 3 minor 15 references

Optimal Dynamical Frames

T0 review · 3 major / 3 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A bounded operator generates a Parseval frame from the iterates of some vectors if and only if it is a contraction and its adjoint is strongly stable; the minimal number of generators is then the defect dimension.

desk verdict A genuinely useful paper with a real fixable gap: the Parseval index formula is right, but the proof needs closure/dimension language instead of false range equalities. read the letter →

arxiv 2506.00567 v3 pith:2P7BL2QT submitted 2025-05-31 math.FA

A. Aguilera , C. Cabrelli , F. Negreira , V. Paternostro This is my paper
classification math.FA MSC 42C1547A1547A4530H10
keywords dynamicalsamplingframetheoryHardyspacesmodelcontractionsParsevalframesindexinnerfunctions
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper settles the existence question for Parseval dynamical frames: an operator $T$ on a separable Hilbert space admits a Parseval frame made of the iterates $\{T^n v_i\}_{i\in I, n\ge 0}$ if and only if $\|T\|\le 1$ and $(T^*)^n v\to 0$ for every $v$. It introduces the frame index, the smallest number of vectors whose orbits form a frame, and proves that in the Parseval case this number is exactly the dimension of the defect space $(I-TT^*)(H)$, attained by linearly independent generators. The general index is expressed as the minimum of those defect dimensions over all contractions similar to $T$. Because the same model-space machinery constructs the frames explicitly, the result turns a qualitative existence statement into a quantitative design principle: how many generators are needed and which ones are optimal. A companion result shows that when both $T$ and $T^*$ admit frames of iterations, their indices coincide.

What carries the argument

The central object is the model-space representation: every frame of iterations is similar (unitarily equivalent, in the Parseval case) to a basic frame $\{A_N^n(P_N e_i)\}_{i\in I,n\ge 0}$ in a model space $N\subseteq H^2_K$, where $A_N$ is the compression of the unilateral shift. The existence and index proofs run through the Rota-de Branges-Rovnyak functional model, which sends $v$ to $\sum_{n\ge 0} D(T^*)^n v\, z^n$ with defect operator $D=(I-TT^*)^{1/2}$ and defect space $K=\overline{D(H)}$. The index formula follows from the identity $I-TT^*=V(I_N-A_N S^*|_N)V^*$ together with $(I_N-A_N S^*|_N)(N)=P_N(K)$. For the adjoint problems, the machinery is Helson's full-range shift-invariant subspaces, inner functions, and the involution $\rho(Q)(z)=Q(\bar z)^*$, which decides when a basic frame and its adjoint frame are similar.

What would settle it

For the diagonal contraction $Te_k=(1-2^{-k})^{1/2}e_k$ on $\ell^2$, the defect operator $D=(I-TT^*)^{1/2}$ has eigenvalues $2^{-k/2}$, so $D(\ell^2)\neq (I-TT^*)(\ell^2)$ as sets; computing the minimal number of Parseval generators for this $T$ directly would settle whether the index formula survives without the false set equality.

Watch

Extended reading notes

Core claim

The paper proves that the previously known sufficient conditions are also necessary: given a bounded operator $T$, there exists a Parseval frame of the form $\{T^n v_i\}_{i\in I,n\ge 0}$ exactly when $T$ is a contraction and its adjoint is strongly stable. It then defines the frame index $\gamma(T)$ and the Parseval frame index $\gamma_p(T)$ and proves $\gamma_p(T)=\dim(I-TT^*)(H)$, with the index attained by a linearly independent set of generators. For general frames, $\gamma(T)=\min\{\dim(I-QQ^*)(H): Q\text{ contraction similar to }T\}$. The paper also shows that if a frame of iterations can be generated by finitely many vectors, every minimal generating set is linearly independent, and that an operator and its adjoint have coinciding indices when both admit frames of iterations, with optimal Parseval frames constructed simultaneously for jointly strongly stable contractions.

Load-bearing premise

The proof of the Parseval index formula relies on the identity $(I-TT^*)^{1/2}(H)=(I-TT^*)(H)$ for a positive operator; that set equality fails in general, so the load-bearing assumption is that the stated dimension is really the dimension of the closure and that the equality can be repaired.

Editorial extensions

If this is right

  • In dynamical sampling terms, the minimum number of spatial sensors is read off the evolution operator: $\gamma_p(T)=\dim(I-TT^*)(H)$.
  • Every operator that admits a Parseval frame of iterations is unitarily equivalent to a shift compression on a model space, so basic frames are a universal normal form for such systems.
  • When a frame of iterations has finitely many generators, every minimal generating set is linearly independent; infinite minimal generating sets can also be chosen linearly independent.
  • For contractions with joint strong stability, optimal Parseval frames for $T$ and $T^*$ exist with the same number of generators and can be built together.
  • A normal operator has Parseval index either $0$ or $\infty$; it can never have a finite number of generators forming a Parseval frame of iterations.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial extension: the explicit index formula suggests a computational recipe for concrete evolution operators: compute the defect operator's range dimension, then pull an orthonormal basis of the defect space back through the model-space map to obtain optimal generators.
  • Editorial caveat: the proof of Theorem 4.2 uses the set equality $(I-TT^*)^{1/2}(H)=(I-TT^*)(H)$ for a positive operator; that equality is false in general, so the formula as stated needs a closure convention or a supplementary argument before the proof is complete.
  • Editorial extension: the joint-construction result for $T$ and $T^*$ points to a duality in sensor placement: optimal generating sets for an evolution and for its adjoint are related by the involution $\rho$ on inner functions, which could be tested numerically on finite-rank models.
  • Editorial extension: for diagonal normal contractions, the paper's examples predict $\gamma_p=\infty$ whenever the defect dimension is infinite; truncating to finite matrices and tracking the lower frame bound as $N$ grows would provide a concrete numerical check of the formula.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 3 minor

Summary. The paper studies frames of the form {T^n v_i}_{i in I, n >= 0} in a separable Hilbert space, called dynamical frames. It proves Theorem 1.1: a Parseval dynamical frame exists if and only if ||T|| <= 1 and (T*)^n v -> 0 for every v. It then introduces the frame index gamma(T) and the Parseval frame index gamma_p(T), claims the formula gamma_p(T) = dim(I - T T*)(H) with attainment by linearly independent generators, and derives a formula for gamma(T) as a minimum over contractions similar to T. The paper also proves that minimal generator sets can be chosen linearly independent, shows that in the Parseval case the dimension of the span of the generators equals gamma_p(T), constructs optimal Parseval frames for jointly strongly stable contractions through model spaces, and analyzes when the frames for T and T* are similar via an involution on inner functions.

Significance. The questions addressed are natural and the overall program is valuable: completing the existence characterization for Parseval dynamical frames and identifying the minimal number of generators has intrinsic operator-theoretic interest and direct relevance to dynamical sampling. The paper makes good use of vector-valued Hardy spaces and model spaces, and it contains several explicit constructions and worked examples, such as Example 4.4 showing 0 < gamma(T) < gamma_p(T) = infinity. The eventual index formula is dimensionally plausible and the cardinal version may well survive the gaps identified below. However, the written proofs of the central index theorem and of the related range identities contain false set equalities and an unjustified closedness assertion. These issues are load-bearing for the paper's main claims, although they appear localized and probably repairable.

major comments (3)
  1. [Section 3, proof of Theorem 3.4] The proof defines the defect space as K = D(H), where D = (I_H - T T*)^(1/2). For a positive operator on an infinite-dimensional Hilbert space, D(H) need not be closed; for example, if D = diag(1/n) on l^2, then D(H) = {y : sum n^2 |y_n|^2 < infinity} is not closed. Since H^2_K is defined only when K is a Hilbert space, the construction of L : H -> H^2_K requires K to be replaced by the closure of D(H), or an explicit argument that D(H) is complete. This is not a cosmetic issue: Theorem 3.4 is the functional-model step behind Theorem 1.1 and behind all later optimal-frame constructions, so the proof as written has a genuine gap.
  2. [Section 4, proof of Theorem 4.2] The proof states: 'Since (I_H - T T*)^(1/2) is self adjoint then (I_H - T T*)^(1/2)(H) = (I_H - T T*)(H).' This set equality is false in general. For a positive operator B, range(B^(1/2)) generally strictly contains range(B); for instance, if B = diag(1/n^2) on l^2, then y = (1/n^2) belongs to range(B^(1/2)) but not to range(B), since the required preimage would be the non-square-summable sequence (1). The upper-bound part of Theorem 4.2 therefore needs a separate argument showing dim range(B^(1/2)) = dim range(B) in the appropriate sense, or a reformulation in terms of closures. The manuscript should also state explicitly whether dim denotes algebraic dimension of the (possibly non-closed) range or dimension of its closure; the current notation is ambiguous and the proof relies on the ambiguity.
  3. [Section 4.2, Lemma 4.10 and Proposition 4.9] Lemma 4.10 asserts without proof that K_N = {g(0) : g in N} is a closed subspace of K, and the subsequent projection argument requires this closedness. The assertion is false in general. In the model space N arising from T = diag(r_n) with r_n = sqrt(1 - 1/n^2) as in the skeptical counterexample, K_N = D(H), which is not closed in l^2. Hence the projection P_{K_N} used in the proof is not defined, and the identity (I_N - A_N S*)(N) = P_N(K) is not established. Consequently Proposition 4.9, which claims gamma_p(T) = dim span{v_i} for an arbitrary Parseval frame, does not follow as stated; it requires either a closure/dimension reformulation or additional hypotheses, and the statement should be corrected or its proof supplied.
minor comments (3)
  1. [Throughout] There are several typographical errors, including 'Preliminares' in the Section 2 heading, 'preservad' instead of 'preserved' in the text preceding Proposition 4.3, and 'aln in N' instead of 'all n in N' in Section 5.2.
  2. [Section 6.2] In the paragraph after equation (9), the displayed frames write T^n(V e_j) and (T*)^n(S* E_j); the first should be T^n(V(P_N e_j)), and the index set should be J rather than I.
  3. [Section 7.1, proof of Theorem 7.5] The proof contains several bracket and notation typos, such as 'langle [rho(Q) f, e_i chi^{-n}rangle' and '[rho(Q) e_j rangle'; these should be corrected to standard inner-product notation before publication.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the Parseval characterization and the frame-index formula are genuinely derived from prior proved results and external classical tools; the flagged false range equalities are proof gaps, not by-construction reductions.

full rationale

Walking the claimed derivation chain, Theorem 1.1 is proved, not assumed: necessity (Proposition 3.3) reduces a Parseval frame of iterations, via Proposition 2.2 (cited to [4]/[12]/[15]), to a unitarily equivalent basic frame in a model space and then computes ||A_N|| <= 1 and strong stability of S*|_N; sufficiency (Proposition 3.5 plus Theorem 3.4) is the Rota/de Branges-Rovnyak functional model (cited to Nikolski [28, p. 18], proof included). Theorem 4.2's formula rests on two independent legs: a lower bound #I >= dim(I_N - A_N S*)(N) = dim(I_H - TT*)(H) obtained from the basic-frame reduction, and an explicit construction whose generator count is dim K_o with K_o = (I_H - TT*)^{1/2}(H). gamma_p(T) is defined as a minimal generating cardinal, not in terms of the defect dimension, so the equality is a genuine theorem rather than a definitional collapse. The only delicate identification is the invoked set identity '(I_H - TT*)^{1/2}(H) = (I_H - TT*)(H)' (proof of Theorem 4.2), which is false for positive operators with non-closed range (e.g., D = diag(1/n) on l^2), and Lemma 4.10's unsupported claim that K_N := {g(0) : g in N} is closed; per the reviewing rule these are flagged explicitly as proof gaps (a correctness risk; the cardinal formula survives under a dimension-of-closures reading), but they are not circular steps, since nothing is defined in terms of the target result and no fitted parameter is renamed as a prediction. Self-citations exist (Propositions 2.1-2.2 from [4]/[12]/[15]; Theorem 1.2 credited to [13]) but their stated assumptions do not include the target results, Theorem 1.2 is re-proved in the paper from Theorem 1.1, and Theorem 3.4 plus Beurling-Lax-Halmos are external (Nikolski; Helson); hence the self-citation is minor and not load-bearing.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The main results rest on standard operator model theory and on the authors' earlier frame-by-orbits results; no free parameters are fitted and no new entities are postulated. The paper is a proof-based contribution, so the ledger lists the external theorems invoked and one unproved sub-lemma.

assumptions (6)
  • domain assumption Proposition 2.2: every frame of iterations is similar to a basic frame {A_N^n(P_N e_i)} in a model space N subset H^2_{ell^2(I)}, and every Parseval frame is unitarily equivalent to such a basic frame.
    Cited from [4, Theorem 3.9], [15, Theorem 3.4], and [12, Theorem 2.1]; this is the bridge that lets the paper reduce arbitrary operators to shift compressions on model spaces.
  • domain assumption Theorem 3.4 (Rota/de Branges-Rovnyak functional model): every contraction with strongly stable adjoint is unitarily equivalent to a shift compression A_N on a model space N subset H^2_K with K = (I - TT*)^{1/2}(H).
    Proved in the paper via the isometry L v = sum D(T*)^n v chi^n; it is the construction underlying Theorem 1.1 and Theorem 4.2.
  • standard math Beurling-Lax-Halmos theorem (Theorem 5.1): S-invariant subspaces of H^2_K are of the form bQ(H^2_{K1}) with Q analytic, and full-range subspaces correspond to inner functions Q.
    Used in Section 5 to characterize joint strong stability via full-range subspaces and in Section 7 for the similarity results.
  • standard math Theorem 5.7 ([29, Theorem 1.2, Chapter 2]): a contraction is strongly stable iff its minimal unitary dilation G satisfies G = direct sum over n in Z of U^n L with L = (U - A_N)N.
    Used to prove Theorem 5.8 that strong stability of A_N implies the orthogonal complement is full range.
  • ad hoc to paper Lemma 4.10: for a model space N, the set K_N = {g(0): g in N} is a closed subspace of K.
    Assumed without proof; the projection P_{K_N} is used in the proof of Lemma 4.10. Likely true for model spaces but no argument is supplied.
  • standard math Nevanlinna and F. Riesz factorization of scalar inner functions, and Proposition 7.9 on constant determinant inner functions in finite dimension.
    Used in Theorem 7.14 and in the finite-dimensional part of Theorem 7.10 to convert inner-function inclusions into divisibility conditions on Blaschke and singular factors.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Optimal Dynamical Frames." pith.science (2026). https://pith.science/paper/2P7BL2QT

@misc{pith2026250600567,
  author       = {Pith},
  title        = {Pith review of: Optimal Dynamical Frames},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2P7BL2QT}},
  note         = {Machine review of arXiv:2506.00567}
}
read the original abstract

Motivated by the dynamical sampling problem, we study frames in an infinite dimensional Hilbert space generated by the iterates of a bounded operator T, also known as dynamical frames. We first characterize the operators that generate Parseval dynamical frames by showing that the previously known sufficient conditions for their existence are also necessary. We then introduce the frame index of T, the minimal number of vectors required to generate a frame by iterations, and derive an explicit formula for it in the Parseval case together with a general condition for the non-Parseval setting. Finally, we prove that if both T and T* admit frames of iterations, then their frame indices coincide through an explicit construction.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

15 extracted references · 15 canonical work pages

  1. [21]

    D ´ıaz Mart´ın, I

    R. D ´ıaz Mart´ın, I. Medri, and U. Molter,Dynamical sampling: a view from control theory, in Excursions in Harmonic Analysis, Vol. 6, Springer, 2021, p. 269–295

    work page 2021

  2. [22]

    D´ıaz Mart´ın, I

    R. D´ıaz Mart´ın, I. Medri, and J. Osorio,Error analysis on the initial state reconstruc- tion problem, Sampling Theory, Signal Processing, and Data Analysis, 21 (2023). Paper No. 18. OPTIMAL DYNAMICAL FRAMES 35

    work page 2023

  3. [23]

    Y. Dong, R. Liu, and H. W ang,Weighted composition operators that preserve p-frames, Journal of Mathematical Analysis and Applications, 539 (2024). Paper No. 128480

    work page 2024

  4. [24]

    R. S. Garcia, J. Mashreghi, and W. T. Ross.Introduction to model spaces and their op- erators.Cambridge Studies in Advanced Mathematics, 148. Cambridge University Press, Cambridge, 2016

    work page 2016

  5. [25]

    Halmos.Shifts on Hilbert spaces.Journal f ¨ur die reine und angewandte Math- ematik, 1961, no

    P. Halmos.Shifts on Hilbert spaces.Journal f ¨ur die reine und angewandte Math- ematik, 1961, no. 208, 102–112, 1961

    work page 1961

  6. [26]

    Helson.Lectures on Invariant Subspaces.Academic Press, London, 1964

    H. Helson.Lectures on Invariant Subspaces.Academic Press, London, 1964

    work page 1964

  7. [27]

    C. S. Kubrusly.Strong stability does not imply similarity to a contraction.Systems Con- trol Lett., 14, 397–400, 1990

    work page 1990

  8. [28]

    N. K. Nikolski.Treatise on the Shift Operator: Spectral function theory.Springer- Verlag, 1986

    work page 1986

Show all 15 references
  1. [29]

    B. Sz. Nagy, C. Foias, H. Bercovici, and L. K ´erchy.Harmonic Analysis of Operators on Hilbert Spaces.Springer, 2010

    2010

  2. [30]

    Paulsen.Every completely polynomially bounded operator is similar to a contraction

    V.I. Paulsen.Every completely polynomially bounded operator is similar to a contraction. J. Funct. Anal. 55, 1–17, 1984

    1984

  3. [31]

    Philipp.Bessel orbits of normal operators.J

    F. Philipp.Bessel orbits of normal operators.J. Math. Anal. Appl., 448, 767–785, 2017

    2017

  4. [32]

    Rosemblum and J

    M. Rosemblum and J. Rovnyak.Hardy Classes and Operator Theory.Oxford Univer- sity Press, New York, 1985

    1985

  5. [33]

    Radjavi and P

    H. Radjavi and P. Rosenthal.Invariant Subspaces.Springer-Verlag, Berlin, 1973

    1973

  6. [34]

    G. C. Rota.On models for linear operators.Comm. Pure Appl. Math. 13, 469–472, 1960

    1960

  7. [35]

    Luis Santal´o

    S. Tang.System identification in dynamical sampling.Adv. Comput. Math., 43, 555– 580, 2017. Departamento de Matem´atica, Universidad de Buenos Aires, Instituto de Matem´atica “Luis Santal´o” (IMAS-CONICET-UBA), Buenos Aires, Argentina. Email address:aaguilera@dm.uba.ar Email a...

    2017

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.