2412.19734
Dynamics, data and reconstruction
Suddhasattwa Das, Tomoharu Suda
correctmedium confidence
- Category
- Not specified
- Journal tier
- Specialist/Solid
- Processed
- Sep 28, 2025, 12:56 AM
- arXiv Links
- Abstract ↗PDF ↗
Audit review
The paper’s Theorem 4 establishes (i) existence of the left/right Kan extensions via completeness/cocompleteness of [DS] and pointwise formulas, and (ii) observation-consistency of the left Kan extension in discrete time, by reducing the Kan-extension colimit to the terminal object in an appropriate left slice using Lemmas 10.2–10.4 (assumption j* left-inverse to j∘ι and a delay trick to reduce jump k to 0) . This matches the intended statement R ∘ Data ∘ ιobs = Frgt ∘ ιobs and yields exact reconstruction for observable sources . By contrast, the model’s proof claims that the object (X0, id) is terminal in the comma category (Data ↓ Y) and that measurement objects canonically identify with j*(1-word alphabets); neither claim is established by the paper’s axioms, and the argument ignores the jump-k issue handled by Lemma 10.3. Hence the model’s conclusion relies on unproven/extraneous assumptions, while the paper’s argument holds as stated.
Referee report (LaTeX)
\textbf{Recommendation:} minor revisions \textbf{Journal Tier:} specialist/solid \textbf{Justification:} The manuscript sets up a useful categorical pipeline from dynamics with observations to time-series data and back, and frames reconstruction as a Kan extension, leading to a crisp observation-consistency theorem in discrete time. The arguments are correct under the stated axioms. Some categorical details of the pointwise colimit/indexing could be made a bit more explicit to improve readability and to reassure non-expert readers, but these are presentational tweaks rather than substantive issues.