feat(longitudinal): consolidate lagged correlation and discreteDRIFTstd - #310
feat(longitudinal): consolidate lagged correlation and discreteDRIFTstd#310seonghobae wants to merge 192 commits into
Conversation
|
Note Reviews pausedIt looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the Use the following commands to manage reviews:
Use the checkboxes below for quick actions:
📝 WalkthroughWalkthrough
Changes종단 시간 표준화와 잔차 로그율 계약
Estimated code review effort: 4 (Complex) | ~60 minutes Merge Risk: 🔵 Low · up to The change is mergeable with owner follow-up for a few bounded issues: clarify that the ln(0.5) example assumes a unit interval, correct the source-access dates, and make mean calculation deterministic with respect to input order. No high-impact merge-blocking risk is identified. Sequence Diagram(s)sequenceDiagram
participant Caller
participant irregular_residual
participant EventTimeInterval
participant LongitudinalError
Caller->>irregular_residual: EventTimedObservation 전달
irregular_residual->>irregular_residual: 단위별 CWC 및 이벤트 시간 정렬
irregular_residual->>EventTimeInterval: 연속 이벤트 간격 생성
EventTimeInterval-->>irregular_residual: 승인된 typed interval 반환
irregular_residual-->>Caller: pairwise-mean log-rate 또는 LongitudinalError 반환
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
Full details: Docstring CoverageExplanation Docstring coverage is 89.62% which is sufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 106 functions across 16 files. (10 skipped: 10 unsupported.)
✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
This comment was marked as resolved.
This comment was marked as resolved.
Sorry, something went wrong.
![devin-ai-integration[bot]](https://avatars.githubusercontent.com/in/811515?s=60&v=4){kind=small}
This comment was marked as resolved.
This comment was marked as resolved.
Sorry, something went wrong.
Record the public +0.0 CWC lag contract, IEEE Std 754-2019 authority, and the Curran/Bauer refusal that this identity repair is not raw-process drift. Meredith (1993) and Mislevy (1991) remain unread.
|
Exact head is now Public CWC lag residuals now use one canonical No self-approval. Independent review is requested on this exact head. Queued/pending/skipped/cancelled/stale/predecessor-head checks are non-passing. Meredith (1993) and Mislevy (1991) remain unread. |
Hamaker Eq. 1a occasion-mean residuals that are exact zero now use one public +0.0 encoding, matching CWC and within/between decomposition.
Record the public +0.0 Hamaker Eq. 1a residual contract, IEEE Std 754-2019 authority, and the refusal that this identity repair is not a within-person or RI-CLPM lag. Meredith (1993) and Mislevy (1991) remain unread.
|
Exact head is now Public Hamaker Eq. 1a occasion-mean lag residuals now use one canonical No self-approval. Independent review is requested on this exact head. Queued/pending/skipped/cancelled/stale/predecessor-head checks are non-passing. Meredith (1993) and Mislevy (1991) remain unread. |
|
@opencode-agent review | Independent exact-current-head formal review requested for |
1 participant
Longitudinal Modeling landing vehicle
This Draft is the coherent Longitudinal Modeling vehicle for event-time lagged Pearson association, stationary/discrete drift-diffusion standardization, within/between decomposition, CWC and irregular-residual composition, occasion-mean event-time composition, known-truth recovery, and the DDD boundary that keeps temporal composition out of
psychometric_coreauthority.Current exact head:
c6680450152b1e0a2c9abb553772d74a23923335against protectedmain@1bc02f580cf48e1d39da239f0e818453437c31c3.Temporal/state composition is owned by Longitudinal Modeling and implemented in
longitudinal_core, with typedEventTimeIntervalretained end-to-end. Reusable static/generalized-mixed/dependence psychometric arithmetic remains fast-mlsirm-owned and may be consumed only through an immutable released contract.Current RED → causal repair: stable Between identity cannot acquire occasion aliases
Fresh review of the GAP-043/GAP-044 recovery-identity repair found that the generic tuple
(unit_index, occasion_index, level)was too permissive forComponentLevel::Between. Longitudinal decomposition defines that component as stable/unit-level and emits exactly one Between component per unit using canonicaloccasion_index = 0. The predecessor RMSE admission nevertheless accepted the same unit-level Between component again under another occasion index, which changed the recovery denominator without adding a new scientific target.Public RED
0a03041c5c1c2152fe90f86204edf2c957cebbfeaddscomponent_rmse_stable_between_identity_contract.rs: one unit supplies two stable Between rows under occasion indices0and2, while a separate Within row carries error. The predecessor accepts both Between aliases as distinct tuple identities and lets the alias weight the RMSE denominator.Review also found that the earlier cross-slice permutation regression itself had encoded nonzero occasion indices on stable Between fixtures. Commit
53563e975dd10e4f18218880cca2bb4cd79d0e73corrects that fixture to canonical Between occasion0while preserving the original transport-permutation test.Causal repair
ec2c1219127e834d89051415f1de8e1cb48507aavalidates bounded-context component identity before either truth or recovered map insertion. StableBetweenrequires canonical occasion index0;Withinretains its actual(unit, occasion)identity. Thus an occasion alias cannot become a second stable target or an implicit weight. Residual arithmetic, deterministic canonical accumulation, overflow-safe scaling, nonzero-RMSE underflow refusal, and exact perfect-recovery zero are unchanged.Research trace
c9cb68d30ea81bf15ec1326e74e94679f80287d5distinguishes the scientific level boundary from TEPP's storage encoding: Curran & Bauer (2011) and Hamaker et al. (2015) support separation of stable between-person information from occasion-specific within-person process;occasion_index = 0is TEPP's canonical carrier for the stable component, not substantive event time. Release fragmentc6680450152b1e0a2c9abb553772d74a23923335records the new admission rule.This remains Longitudinal Modeling / Validation Evidence identity and projection policy, not reusable static psychometric arithmetic. A separately weighted recovery metric needs an explicit weight contract; duplicate rows and Between occasion aliases are never implicit weights. No fast-mlsirm source copy or mutable sibling dependency was introduced.
Retained scientific lineage
698f12f5...→2fae4cb2...), and a mathematically nonzero component RMSE cannot underflow into false perfect recovery (496583c6...→a82b383b...).f64result (8ad72ac9...→2dd9537e...;5fb93c40...→025dce7f...; edge976ce7d7...).ln; RED766ddc7a...and repair16f21d9a...use exact represented difference/ln_1pwithin the factor-two neighborhood while retaining direct-ratio/log-domain fallbacks at extreme scales.96f1c334...→ae5081d8...).+0.0identity only where zero has no directional scientific meaning; private numerical intermediates are not blanket-canonicalized.±1; strict-interior covariance cannot round into a false perfect endpoint, and nonzero covariance cannot round into false zero.discreteDRIFTstdanddiscreteDIFFUSIONstdretain their endpoint/subnormal refusal contracts.#486 verified-successor consolidation
Commit
db335d904ccc43dd0af7982c34df83008c968ef6preserves #486's valid operational release delta asCHANGELOG.d/longitudinal-occasion-mean-event-time.mdunder the correctlongitudinal_coreowner. The survivor also carries Hamaker Eq. 1a scientific evidence, signed-zero-normalized numeric occasion identity, duplicate/sparse refusal, the shared stable mean, typed finite-positive event intervals, exact irregular-residual log-rate reuse, and explicit refusal to reinterpret the result as a within-person/RI-CLPM lag. Wrong-owner temporal source, raw-bit signed-zero identity, naivesum/count, mixed timezone provenance, and unsupported operational claims were intentionally not inherited.Curran, P. J., & Bauer, D. J. (2011). The disaggregation of within-person and between-person effects in longitudinal models of change. Annual Review of Psychology, 62, 583–619. https://doi.org/10.1146/annurev.psych.093008.100356
Hamaker, E. L., Kuiper, R. M., & Grasman, R. P. P. P. (2015). A critique of the cross-lagged panel model. Psychological Methods, 20(1), 102–116. https://doi.org/10.1037/a0038889
Exact-head merge bar
Exact head
c6680450152b1e0a2c9abb553772d74a23923335remains Draft. No predecessor check or review evidence transfers after the RED/fixture-repair/source-repair/trace/changelog commits. Fresh exact-head Rust/documentation/security workflows and a qualifying independent current-headAPPROVEDsubmission are required. Live branch/ruleset requirements remain authoritative. Do not self-approve, force-push, destructively rebase, weaken coverage/security gates, or use admin bypass as routine delivery evidence. Merge only after every required exact-head gate is satisfied.