{"id":"4267ded6-7b32-4f4f-bc48-9c225fc78793","arxiv_id":"2509.04531","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"OrbDot is a Python package that fits secular evolution models to combined transit, eclipse, and radial velocity data using nested sampling, with automated interpretation reports.","lead":"OrbDot is a new Python package for detecting slow changes in exoplanet orbits by fitting models to transit, eclipse, and radial velocity data together. It uses nested sampling for Bayesian inference and includes tools for interpretation, aiming to make long-term orbital evolution studies more accessible.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No in-paper validation of OrbDot's secular models or sampling: the central reliability claim rests on external examples, so a bug in the likelihood/prior implementation would not be caught.","rationale":"I agree with the reader's CONDITIONAL verdict. The weakest assumption is not that nestle/PyMultiNest are unreliable per se—they are established—but that OrbDot's likelihood functions, priors, and parameter transformations correctly turn physical model parameters into predicted timings/RVs and that the wrapper's evidence estimates are trustworthy. The paper's support is entirely external: the WASP-12 b example, TrES-1 b case study, and LTT-9779 b application are not reproduced in the preprint. For a software paper, an in-paper validation (synthetic recovery or a unit-test report) is the minimal evidence needed. Absent that, the central claim is plausible but unverified. The missing direct code URL compounds this but is secondary. I therefore leave the reader's CONDITIONAL verdict unchanged.","tokens_in":3869,"tokens_out":3716,"duration_ms":37227,"concrete_test":"Run a synthetic injection-recovery benchmark with the released OrbDot package: simulate 100 realizations of transit/eclipse mid-times and RVs from a known secular model (e.g., dP/dt = -1e-10, apsidal precession), fit with OrbDot's nested sampling, and verify that 1D posteriors cover injected values at the expected frequency and that logZ for the true model beats a constant-period model. Also run `pytest` on the package's test suite and compare the WASP-12 b demo output with the published decay rate. If injected signals are not recovered, the central reliability claim fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"OrbDot's central claim is that it can reliably fit secular evolution models to transit/eclipse times, durations, and RVs. For this to hold, its likelihood and prior implementations must correctly map physical parameters (dP/dt, precession rate, RV semi-amplitude, etc.) to observables, and the nested-sampling wrapper must give unbiased posteriors/evidences. The preprint offers no in-paper test, synthetic recovery, or benchmark of either. It says 'examples demonstrate that OrbDot can quickly reproduce literature results' and cites Hagey et al. (2025) and Edwards et al. (2023), but these are external works; a reader cannot check from the paper whether WASP-12 b's decay rate, TrES-1 b's parameters, or LTT-9779 b's fit were recovered with correct uncertainties. This matters because secular signals are tiny (e.g., dP/dt ~ 1e-10 d/d) and easy to miss or overstate if the timing model or prior volume is wrong. A sign error or missing term in, say, the transit-duration model would not be exposed by the paper's narrative. Thus the load-bearing assumption—model correctness and code correctness—is currently unsupported by the preprint itself. This is not an allegation of error; it is a request for evidence.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"OrbDot is presented as a Python package for fitting secular orbital evolution models to exoplanet transit/eclipse mid-times, transit durations, and radial velocities using nested sampling. The paper describes the package's intended capabilities: simultaneous multi-dataset fitting, model/prior flexibility, Bayesian evidence computation, and report generation for physical interpretation. It argues that no existing tool provides the full suite of functionality and that OrbDot lowers the barrier for studying long-term orbital evolution. The authors cite companion papers (Hagey et al. 2025; Edwards et al. 2023) and claim that online examples reproduce literature results, but the manuscript itself contains no code listings, no equations, no test results, no benchmark figures, and no quantitative validation of the implemented models or sampling workflows.","tokens_in":4142,"tokens_out":2319,"duration_ms":27571,"significance":"If the software functions as claimed and is indeed validated by the external examples, it would fill a genuine niche: a single, maintained, open-source tool for secular-evolution inference from heterogeneous timing and RV data, with Bayesian model selection built in. The authors' prior applications to WASP-12 b, TrES-1 b, and LTT-9779 b suggest real-world utility. However, the current manuscript provides no in-paper evidence that the likelihoods, priors, and nested-sampling wrappers are correct. Because the entire claim is that the package reliably fits secular models, the absence of validation is a load-bearing gap. The paper is also unusually spare for a software paper: it does not give a repository URL, version number, API examples, or a description of the secular models implemented. The write-up is more a proposal than a demonstration.","major_comments":[{"comment":"The paper states 'The examples demonstrate that OrbDot can quickly reproduce literature results using only a few lines of code,' but no examples, outputs, or quantitative comparisons are shown in the manuscript. The cited Hagey et al. (2025) and Edwards et al. (2023) are external publications that (a) are not part of this paper and (b) need not have used exactly the same model code, priors, or data reductions as OrbDot. As written, the central reliability claim—that OrbDot correctly and robustly fits secular evolution models—is unsupported. Please add at least one in-paper validation: synthetic data injection with known dP/dt and precession parameters, recovery of literature values for a benchmark system (e.g., WASP-12 b) with quoted uncertainties, or a comparison against an independent sampler on a simple test model. This is not an accusation of error; it is a request for the evidence t","section":"Summary, fourth paragraph"},{"comment":"No equations or explicit model definitions are provided for the secular evolutions (orbital decay, apsidal precession, proper-motion terms, companion-induced effects). A reader cannot assess whether the mapping from physical parameters (dP/dt, precession rate, RV semi-amplitude, etc.) to observables is implemented correctly or what priors are available. The paper should include a 'Models' section listing the mathematical forms fitted, or at least give explicit pointers to equations in the companion papers and documentation. Without this, the manuscript cannot support the claim that the software 'unifies transit, eclipse, and RV data' in a scientifically meaningful way.","section":"Summary / Statement of need"},{"comment":"The comparison with existing codes would be more persuasive if it were grounded in quantitative capabilities. For example, the claim that PdotQuest 'currently supports only the orbital decay model' and that Susie uses 'simple least-squares fitting' needs citation to the specific code versions tested. More importantly, the absence of any benchmark timing or runtime comparison makes the 'highly efficient' claim in the Summary an assertion rather than a demonstrated property. Please provide representative runtimes or an efficiency comparison for a standard model (e.g., WASP-12 b) if efficiency is part of the contribution.","section":"Similar software"}],"minor_comments":[{"comment":"No repository URL, DOI, license, or version number is given. For a software paper, this is essential. At minimum add the GitHub/ReadTheDocs links and a license statement.","section":"Throughout"},{"comment":"The phrase 'measurements of secular variations can even probe the interior structure of exoplanets' is intriguing but not explained. A one-sentence physical justification (e.g., tidal Love numbers from apsidal precession) would help.","section":"Summary, first paragraph"},{"comment":"The claim that 'OrbDot... is presented to the community with transparency that lends itself to community contributions and independent verification' is not yet evidenced; the manuscript contains no contribution guidelines or test suite description.","section":"Statement of need"},{"comment":"Minor typographical issue: 'OrbDot remains highly efficient' in the Summary has an odd line break ('thenestle'), and the text alternates between 'OrbDot' and 'OrbDot.' with inconsistent spacing. A copyedit is needed.","section":"Similar software"}],"recommendation":"major_revision","confidential_remarks":"I am sympathetic to the authors' approach: they have clearly done applied work with earlier versions of this code, and a well-packaged secular-evolution fitting tool would benefit the community. The problem is that this submission is a software paper that does not demonstrate the software. I would not reject outright, because the missing validation and model definitions are fixable additions rather than fundamental errors. But the manuscript in its current form is more of an announcement than a refereed software publication. The authors should be asked to include a minimally reproducible benchmark (synthetic or real) and to expose the mathematical model in the paper itself. If they are unwilling to do so, rejection would be appropriate."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"OrbDot is a legitimate entry in a niche that hasn't had a single tool: fitting secular orbital evolution models (decay, precession, proper motion, companions) to transit/eclipse times, durations, and RVs together, using nested sampling. The package has already been used in published work (TrES-1 b, LTT-9779 b, WASP-12 b example), so it isn't vaporware. The comparison to Susie, PdotQuest, TTVFast, and the general-purpose fitters is fair, and the claim that none of them covers this exact combination is probably true as of now.\n\nWhat it does well: the authors have thought about usability — simple input files, interpretation reports, model comparison tools — and they are explicit that it isn't a black box. The documentation and examples on ReadTheDocs are the right idea.\n\nThe soft spot is exactly the stress-test note: the paper ships no validation. There is not a single test, synthetic recovery, benchmark, or error bar in the preprint. The reader is asked to trust that the likelihood mapping and priors are correct because a couple of companion papers used an earlier version. For a software paper, that is a real gap, but it is also fixable. I would not call it fatal; many good JOSS and open-source papers do the same and the code itself can carry the tests. But the paper's own claim to be 'reliable' goes beyond what is shown here.\n\nTwo other things. The phrase 'first software' is likely to age badly and is unnecessary; there are other tools emerging in this space. And the manuscript never gives a direct repository URL — the reader has to go to ReadTheDocs to find it. That is an odd miss for a software announcement.\n\nMy bottom line: the physics is established, the niche is real, and the authors have genuine skin in the game. The paper deserves referee time, and a good referee should ask for one in-paper synthetic recovery (e.g., inject a dv/dt signal, show the sampler recovers it) and a pointer to the repo and test suite. I would not desk-reject this, and I would not send it back for a rewrite; it needs a modest validation add-on.","headline":"OrbDot fills a real gap with a serious tool, but the paper is nearly empty of in-paper evidence that the code actually does what it says; a minimal benchmark would settle it.","tokens_in":4630,"tokens_out":2982,"would_cite":true,"duration_ms":29960,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"OrbDot claims to be the first fully packaged tool for fitting decade-long exoplanet orbit changes to transit, eclipse, and radial-velocity data at once.","keywords":["secular evolution","exoplanet orbits","orbital decay","apsidal precession","transit timing variations","radial velocities","nested sampling","open-source astronomy software"],"falsifier":"Run the paper's advertised WASP-12 b example and compare the fitted orbital-decay rate with the value in the independent literature it claims to reproduce; a disagreement beyond quoted uncertainties would refute the usability claim. Stronger still is an injection-recovery test: generate synthetic transit and RV timings with known decay and precession rates, fit them with OrbDot, and check that the true parameters fall inside the reported credible intervals and that the correct model wins the evidence comparison.","tokens_in":3752,"feed_emoji":"🪐","tokens_out":7569,"duration_ms":67993,"temperature":0.7,"pith_summary":"OrbDot is a Python software package built to detect and characterize the slow, secular changes in exoplanet orbits—orbital decay, apsidal precession, and companion-induced effects—that become visible only after a decade or more of observations. Its central claim is that it is the first fully packaged, documented tool that fits all of the relevant data types at once: transit mid-times, eclipse mid-times, transit durations, and radial velocities. Instead of fitting light curves, it works directly from timing measurements, and it uses nested sampling for Bayesian parameter estimation and model comparison. If the claim holds, researchers can test competing explanations for observed timing anomalies—tidal dissipation, precession, proper-motion effects, or unseen companions—in a few lines of code, and the approach could eventually probe the interior structure of exoplanets through their tidal response. The paper also states the package is efficient enough to handle the growing archive of multi-decade exoplanet observations.","feed_headline":"One sampler fits orbital decay and precession across all timing data","feed_subtitle":"OrbDot unifies transit, eclipse, and RV timings in one Bayesian tool built to catch decade-scale orbit changes.","key_machinery":"The load-bearing mechanism is a nested-sampling Bayesian engine wrapped around a library of long-term orbital evolution models. Nested sampling—an inference method that produces both parameter posteriors and Bayesian evidence values—is what lets OrbDot compare competing physical models, such as tidal decay versus apsidal precession versus companion-induced effects, rather than reporting a single best fit. The models act directly on transit and eclipse mid-times, transit durations, and radial velocities, so heterogeneous data sets can be combined in one fit, which the paper argues helps break parameter degeneracies.","core_discovery":"The paper's claim is that OrbDot is a complete, maintained software suite for secular orbital evolution studies, uniting transit, eclipse, and radial-velocity data in one framework that skips light-curve fitting and fits evolutionary models directly to timing measurements. It carries out Bayesian inference with nested sampling, with freedom to choose models, priors, and parameterizations, and adds an interpretation layer that reports model comparisons, tidal decay parameters, predicted precession rates, and companion dynamical effects. The authors assert no comparable open-source tool combines these capabilities: the most similar codes fit secular models only to timing data, general-purpose","pith_inferences":["If the timing-only design is as reliable as claimed, OrbDot could serve as a community-standard cross-check for newly published secular-trend detections, since its independent re-analysis of the same mid-times would either confirm or challenge decay claims.","A natural stress test the paper does not report is an injection-recovery benchmark: simulate timing data with known decay and precession rates, and check whether OrbDot's posteriors and evidence values recover the truth; its reliability ultimately rests on such a test rather than on the three application examples.","The framework's separation of likelihood from sampler means new secular models could be added without touching the inference layer, a testable design promise that would make the package a platform for studying effects like general-relativistic precession or Kozai–Lidov oscillations.","Because the package reports model comparisons through Bayesian evidence, users' conclusions inherit the well-known sensitivity of nested-sampling evidence to prior choices—something the paper does not discuss."],"forward_implications":["A researcher with only a list of free parameters and a chosen model can fit a decade of heterogeneous observations; the paper demonstrates the workflow by reproducing the literature result of WASP-12 b's orbital decay with a few lines of code.","Simultaneously fitting transit mid-times, eclipse mid-times, and radial velocities breaks parameter degeneracies that single-data-type analyses cannot resolve.","Bayesian model comparison becomes routine, so a reported timing anomaly can be checked against tidal decay, apsidal precession, proper-motion, and companion hypotheses within one framework.","The interpretation reports connect fitted parameters directly to physical theory, such as tidal energy dissipation parameters and predicted precession rates, including implications for non-resonant companions.","Because the nested-sampling machinery handles many free parameters efficiently, the package scales to the growing TESS-era archive of multi-decade ephemerides."],"supporting_citations":[{"why":"Foundational nested-sampling algorithm; supplies the Bayesian evidence and posterior machinery on which OrbDot's inference is built.","marker":"Skilling, 2006"},{"why":"The nestle package implementing nested sampling; one of the two samplers OrbDot wraps.","marker":"Barbary, 2021"},{"why":"Presents PyMultiNest, the second sampling engine OrbDot uses for evidence estimation.","marker":"Buchner et al., 2014"},{"why":"MULTINEST algorithm that PyMultiNest implements; the underlying sampler behind OrbDot's high-dimensional inference.","marker":"Feroz et al., 2009"},{"why":"TrES-1 b case study said to illustrate OrbDot's full capabilities in scientific context; serves as the paper's primary validation example.","marker":"Hagey et al., 2025"},{"why":"Published LTT-9779 b analysis that used an early version of the code; establishes prior scientific use.","marker":"Edwards et al., 2023"},{"why":"Susie; the most similar existing code, used to argue the gap OrbDot fills.","marker":"Barker et al., 2024"},{"why":"PdotQuest; similar timing-only tool restricted to the orbital decay model, contrasted with OrbDot's broader scope.","marker":"Wang et al., 2024"}],"fun_headline_variants":["OrbDot: one Bayesian framework for secular orbit evolution","Nested sampling fits orbit decay and precession from all timings","Transit, eclipse, RV: OrbDot unifies timing data for orbit models","Python tool models exoplanet secular evolution without light curves","OrbDot: from timing data to tidal decay and precession rates"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The load-bearing premise is that the secular evolution models are physically correct and numerically implemented without bugs, and that the nested sampling engines return trustworthy posterior and evidence estimates; the paper offers no in-paper benchmark of its own to verify this.","fun_headline_variants_meta":{"raw":{"variants":["OrbDot: one Bayesian framework for secular orbit evolution","Nested sampling fits orbit decay and precession from all timings","Transit, eclipse, RV: OrbDot unifies timing data for orbit models","Python tool models exoplanet secular evolution without light curves","OrbDot: from timing data to tidal decay and precession rates"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000177,"raw_usage":{"total_tokens":1044,"prompt_tokens":575,"completion_tokens":469,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":319,"completion_tokens_details":{"reasoning_tokens":392}},"tokens_in":319,"tokens_out":469,"duration_ms":4436,"temperature":1.0,"reasoning_tokens":392,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T10:40:39.861083+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the paper's advertised WASP-12 b example and compare the fitted orbital-decay rate with the value in the independent literature it claims to reproduce; a disagreement beyond quoted uncertainties would refute the usability claim. Stronger still is an injection-recovery test: generate synthetic transit and RV timings with known decay and precession rates, fit them with OrbDot, and check that the true parameters fall inside the reported credible intervals and that the correct model wins the evidence comparison.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The nestle package implementing nested sampling; one of the two samplers OrbDot wraps."},{"cited_title":"R., Edwards, B., Tsiaras, A., Boley, A","cited_arxiv_id":null,"evidence_quote":"TrES-1 b case study said to illustrate OrbDot's full capabilities in scientific context; serves as the paper's primary validation example."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Susie; the most similar existing code, used to argue the gap OrbDot fills."}],"review_version":1}