diff --git a/CHANGELOG.md b/CHANGELOG.md index 062a69412..0c2daa0c7 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -40,6 +40,7 @@ All notable changes to TEPP are documented here. The format follows Keep a Chang - `event_core` adds bounded Allen interval-consistency classification, atomic path-consistency closure, contradiction/resource refusals, and an explicit dependency-error fallback without claiming unrestricted global satisfiability. +- `psychometric_core` recovers the Driver, Oud, and Voelkle (2017, p. 16 `DIFFUSIONstd`; Eq. 4, p. 5; Table 2, p. 12; footnote 4; 2017-era ctsem `summary.ctsemFit.R`; JSS PDF re-opened 2026-09-01T19:20Z from https://www.jstatsoft.org/index.php/jss/article/download/v077i05/1104) scalar standardised continuous diffusion on current main after `0ce16e8` dropped the pre-consolidation code while research notes already named the map (register items 59–60). Page 16 prints standardised matrices with the suffix `std` when appropriate. The printed example on p. 16 is `discreteDRIFTstd`, not `DIFFUSIONstd`. Footnote 4 standardises using only the relevant variance, not the total. Table 2 names `DIFFUSION` `Q` the lower-triangular `n.latent × n.latent` Cholesky of diffusion variance/covariance. Process noise is within-subject stochastic input, so that relevant variance is within-subject `asymDIFFUSION` `p = −q / (2 a)`. The 2017-era source forms unstandardised `DIFFUSION` as `mxEval(DIFFUSION, mxobj, compute=TRUE)` and forms `asymDIFFUSIONstd` when `verbose = TRUE`. That source does not form a `DIFFUSIONstd` matrix; the scalar map is the footnote 4 standardisation of that named continuous diffusion: `q / p` after strictly positive `p`. Form strictly positive `p` first, then divide `q` by `p`. In the scalar stationary case that ratio equals `−2 a` and does not depend on `q` once `q > 0`. Unstandardised `q` is defined for growing `a ≥ 0` and for a zero process; standardised `DIFFUSION` is not. Zero `q` has no positive process SD and fails closed. Lasting `p` requires stable `a < 0`. A non-event clock fails closed. `Q_Δt / p` is `discreteDIFFUSIONstd` and depends on the event interval. `p / p = 1` is `asymDIFFUSIONstd` and recovers the same number when `a = −0.5` and remains a distinct named quantity. `q / (trait + p + added)` uses the total and is not this map when `TRAITVAR` is nonzero. Meredith (1993) remains unread (Unpaywall 2026-08-26T17:20Z: `is_oa: false`; OpenAlex closed; Springer `content/pdf` is an HTML stub). Mislevy (1991, *Psychometrika, 56*, 177–196) remains unread on the same terms (DOI `10.1007/bf02294457`; Unpaywall `is_oa: false`). Still not a Kalman filter, not a matrix `expm`, not ESEM estimation, not DSEM, and not ctsem estimation. - `psychometric_core` recovers the Driver, Oud, and Voelkle (2017, Table 2, p. 12 `MANIFESTTRAITVAR`; §7.1, p. 19; p. 16 `MANIFESTTRAITVARstd`; footnote 4; 2017-era ctsem `summary.ctsemFit.R`; JSS PDF re-opened 2026-08-27T14:20Z from https://www.jstatsoft.org/index.php/jss/article/download/v077i05/1104) scalar standardised manifest-trait variance on current main after `0ce16e8` dropped the pre-consolidation code while research notes already named the map (register items 83–84). Table 2 names `MANIFESTTRAITVAR` `Ψ_τ` the additional time-invariant variance-covariance on the measurement level and sets it `NULL` when there is no manifest trait. Equation 5 writes `Γ ~ N(τ, Ψ)` and names that covariance the manifest traits. Section 7.1 names manifest traits stable individual differences in indicator levels, distinct from process-level `TRAITVAR` `φ_ξ`. Page 16 prints standardised matrices with the suffix `std` when appropriate. The printed example on p. 16 is `discreteDRIFTstd`, not `MANIFESTTRAITVARstd`. Footnote 4 standardises using only the relevant variance, not the total. The relevant variance for that named indicator-level correlation is `MANIFESTTRAITVAR`, not process-level `TRAITVAR` and not residual `MANIFESTVAR` `θ`. The 2017-era source forms `MANIFESTTRAITVARstd` only when `MANIFESTTRAITVAR != 0`, as `solve(sqrt(diag(MANIFESTTRAITVAR) + ridging)) %&% MANIFESTTRAITVAR` when `verbose = TRUE`. OpenMx `%&%` is `t(A) %*% B %*% A`. Unlike `TRAITVARstd`, that formation adds `diag(c(ridging), n.manifest)`. The default `ridging = FALSE` adds 0, not `0.0001`; that ridge is a numerical hack and is not this exact map. The scalar correlation is `ψ / ψ = 1` after strictly positive `MANIFESTTRAITVAR`. Form strictly positive `ψ` first, then `1 / √ψ`, then `(1 / √ψ) ψ (1 / √ψ)`. Unstandardised `MANIFESTTRAITVAR` is defined for a zero trait; standardised `MANIFESTTRAITVAR` is not. Zero `MANIFESTTRAITVAR` skips forming `MANIFESTTRAITVARstd` in the 2017-era source and fails closed here. Indicator-level trait variance is an event-time structural quantity, so a non-event clock fails closed. `MANIFESTTRAITVAR` does not require stable `a < 0`. Distinct positive `ψ` recover the same 1. `trait / trait = 1` is `TRAITVARstd` and recovers the same number and remains a distinct named quantity. `θ` is `MANIFESTVAR` and is measurement error, not this correlation. Meredith (1993) remains unread (web search 2026-08-27T14:20Z: Springer/Cambridge Core paywalled; Unpaywall historically `is_oa: false`; Springer `content/pdf` is an HTML stub). Mislevy (1991, *Psychometrika, 56*, 177–196) remains unread on the same terms (DOI `10.1007/bf02294457`). Still not a Kalman filter, not a matrix `expm`, not ESEM estimation, not DSEM, and not ctsem estimation. - `psychometric_core` recovers the Driver, Oud, and Voelkle (2017, Table 2, p. 12 `TRAITVAR`; §7.1, pp. 18–19; p. 16 `TRAITVARstd`; footnote 4; 2017-era ctsem `summary.ctsemFit.R`; JSS PDF re-opened 2026-08-26T17:45Z from https://www.jstatsoft.org/index.php/jss/article/download/v077i05/1104) scalar standardised trait variance on current main after `0ce16e8` dropped the pre-consolidation code while research notes already named the map (register items 81–82). Table 2 names `TRAITVAR` `φ_ξ` the latent trait variance/covariance and sets it `NULL` when there is no trait. Section 7.1 names traits the stable between-subject differences (unit-level unobserved heterogeneity). Page 16 prints standardised matrices with the suffix `std` when appropriate. The printed example on p. 16 is `discreteDRIFTstd`, not `TRAITVARstd`. Footnote 4 standardises using only the relevant variance, not the total. The relevant variance for that named between-subject correlation is `TRAITVAR`, not free first-occasion `T0VAR` and not process-dynamics `asymDIFFUSION`. The 2017-era source forms `TRAITVARstd` only when `TRAITVAR != 0`, as `solve(sqrt(diag(TRAITVAR))) %&% TRAITVAR` when `verbose = TRUE`. OpenMx `%&%` is `t(A) %*% B %*% A`. Unlike `T0VARstd`, that formation uses `diag(diag(TRAITVAR))` and does not add `diag(c(ridging))`. The ridge is a `T0VAR` numerical hack and is not this exact map. The scalar correlation is `trait / trait = 1` after strictly positive `TRAITVAR`. Form strictly positive `trait` first, then `1 / √trait`, then `(1 / √trait) trait (1 / √trait)`. Unstandardised `TRAITVAR` is defined for a zero trait; standardised `TRAITVAR` is not. Zero `TRAITVAR` skips forming `TRAITVARstd` in the 2017-era source and fails closed here. Between-subject variance is an event-time structural quantity, so a non-event clock fails closed. `TRAITVAR` does not require stable `a < 0`. Distinct positive `trait` recover the same 1. `p_0 / p_0 = 1` is `T0VARstd` and recovers the same number and remains a distinct named quantity. `t0_b² v` is `addedT0TIPREDVAR` and is extra first-occasion TI variance, not this correlation. Meredith (1993) remains unread (Unpaywall 2026-08-26T17:20Z: `is_oa: false`; OpenAlex closed; Springer `content/pdf` is an HTML stub). Mislevy (1991, *Psychometrika, 56*, 177–196) remains unread on the same terms (DOI `10.1007/bf02294457`; Unpaywall `is_oa: false`). Still not a Kalman filter, not a matrix `expm`, not ESEM estimation, not DSEM, and not ctsem estimation. diff --git a/crates/psychometric_core/src/error.rs b/crates/psychometric_core/src/error.rs index 4ab2695e0..1ffe2292a 100644 --- a/crates/psychometric_core/src/error.rs +++ b/crates/psychometric_core/src/error.rs @@ -605,6 +605,23 @@ pub enum PsychometricError { /// `asymDIFFUSIONstd`. The continuous-diffusion ratio is not /// the correlation form of `asymDIFFUSION`. StandardisedContinuousDiffusionIsNotStandardisedAsymptoticDiffusion, + /// Driver p. 16 `DIFFUSIONstd` was requested without a strictly + /// positive `asymDIFFUSION`. Footnote 4 standardises using only + /// the relevant variance; zero `q` has no positive process SD. + StandardisedContinuousDiffusionRequiresPositiveStationaryVariance, + /// Driver Table 2 unstandardised `DIFFUSION` `q` was treated as + /// `DIFFUSIONstd`. Unstandardised diffusion is defined for + /// growing `a ≥ 0` and for zero diffusion; standardised + /// `DIFFUSION` is not. + UnstandardisedContinuousDiffusionIsNotStandardisedContinuousDiffusion, + /// Driver p. 16 `discreteDIFFUSIONstd` `Q_Δt / p` was treated + /// as `DIFFUSIONstd`. The discrete map depends on the event + /// interval; the continuous ratio does not. + DiscreteStandardisedDiffusionIsNotStandardisedContinuousDiffusion, + /// Driver §7.1 total `q / (trait + p + added)` was treated as + /// p. 16 `DIFFUSIONstd`. Footnote 4 uses only within-subject + /// `asymDIFFUSION`, not the total. + TotalVarianceScaledDiffusionIsNotStandardisedContinuousDiffusion, /// Driver p. 16 `TIPREDVARstd` was treated as p. 16 /// `asymDIFFUSIONstd`. Equal numbers of 1 after a strictly /// positive relevant variance are still distinct named @@ -1168,6 +1185,18 @@ impl fmt::Display for PsychometricError { Self::StandardisedContinuousDiffusionIsNotStandardisedAsymptoticDiffusion => { "standardised continuous diffusion is not standardised asymptotic diffusion" } + Self::StandardisedContinuousDiffusionRequiresPositiveStationaryVariance => { + "standardised continuous diffusion requires strictly positive stationary within-subject variance" + } + Self::UnstandardisedContinuousDiffusionIsNotStandardisedContinuousDiffusion => { + "unstandardised continuous diffusion is not standardised continuous diffusion" + } + Self::DiscreteStandardisedDiffusionIsNotStandardisedContinuousDiffusion => { + "discrete standardised diffusion is not standardised continuous diffusion" + } + Self::TotalVarianceScaledDiffusionIsNotStandardisedContinuousDiffusion => { + "total-variance scaled diffusion is not standardised continuous diffusion" + } Self::StandardisedTimeIndependentPredictorVarianceIsNotStandardisedAsymptoticDiffusion => { "standardised time-independent predictor variance is not standardised asymptotic diffusion" } @@ -1979,6 +2008,30 @@ mod tests { ); } + #[test] + fn standardised_continuous_diffusion_boundary_messages_are_stable() { + assert_eq!( + PsychometricError::StandardisedContinuousDiffusionRequiresPositiveStationaryVariance + .to_string(), + "standardised continuous diffusion requires strictly positive stationary within-subject variance" + ); + assert_eq!( + PsychometricError::UnstandardisedContinuousDiffusionIsNotStandardisedContinuousDiffusion + .to_string(), + "unstandardised continuous diffusion is not standardised continuous diffusion" + ); + assert_eq!( + PsychometricError::DiscreteStandardisedDiffusionIsNotStandardisedContinuousDiffusion + .to_string(), + "discrete standardised diffusion is not standardised continuous diffusion" + ); + assert_eq!( + PsychometricError::TotalVarianceScaledDiffusionIsNotStandardisedContinuousDiffusion + .to_string(), + "total-variance scaled diffusion is not standardised continuous diffusion" + ); + } + #[test] fn standardised_trait_variance_boundary_messages_are_stable() { assert_eq!( diff --git a/crates/psychometric_core/src/event_time.rs b/crates/psychometric_core/src/event_time.rs index a29bc5c18..213d4f300 100644 --- a/crates/psychometric_core/src/event_time.rs +++ b/crates/psychometric_core/src/event_time.rs @@ -1930,10 +1930,11 @@ pub fn refuse_standardised_asymptotic_diffusion_as_standardised_initial_latent_v /// and remains a distinct named quantity. `DIFFUSIONstd` /// `q / p = −2 a` is the continuous-diffusion ratio and is not this /// correlation. `TIPREDVARstd` `v / v = 1` recovers the same number -/// and remains a distinct named quantity. This crate does not -/// currently export `DIFFUSIONstd` or `TIPREDVARstd`; the refuse -/// still names those quantities. This is not a Kalman filter, not a -/// matrix `expm`, not DSEM, and not ctsem estimation. +/// and remains a distinct named quantity. This crate now exports +/// `DIFFUSIONstd`. This crate does not currently export +/// `TIPREDVARstd`; the refuse still names that quantity. This is +/// not a Kalman filter, not a matrix `expm`, not DSEM, and not +/// ctsem estimation. /// Exact scalar p. 16 `TRAITVARstd` after strictly positive `TRAITVAR`. /// /// Driver, Oud, and Voelkle (2017, Table 2, p. 12; §7.1, pp. 18–19; @@ -2106,9 +2107,8 @@ pub fn refuse_standardised_initial_latent_variance_as_standardised_asymptotic_di /// /// `q / p = −2 a` is the continuous-diffusion ratio. `asymDIFFUSIONstd` /// is the correlation form of `asymDIFFUSION`. Equal numbers when -/// `a = −0.5` remain distinct named quantities. This crate does not -/// currently export `DIFFUSIONstd`; the refuse still names that -/// quantity. +/// `a = −0.5` remain distinct named quantities. This crate now +/// exports `DIFFUSIONstd`. /// /// # Errors /// @@ -2125,6 +2125,130 @@ pub fn refuse_standardised_continuous_diffusion_as_standardised_asymptotic_diffu Err(PsychometricError::StandardisedContinuousDiffusionIsNotStandardisedAsymptoticDiffusion) } +/// Exact scalar p. 16 `DIFFUSIONstd` after strictly positive +/// `asymDIFFUSION`. +/// +/// Driver, Oud, and Voelkle (2017, Table 2, p. 12; Eq. 4, p. 5; +/// p. 16; footnote 4; 2017-era ctsem `summary.ctsemFit.R`; JSS PDF +/// re-opened 2026-09-01T19:20Z from +/// ) +/// name `DIFFUSION` `Q` the lower-triangular `n.latent × n.latent` +/// Cholesky of diffusion variance/covariance. Page 16 prints +/// standardised matrices with the suffix `std` when appropriate. +/// The printed example on p. 16 is `discreteDRIFTstd`, not +/// `DIFFUSIONstd`. Footnote 4: standardisations use only the +/// relevant variance, not the total. `DIFFUSION` is the process +/// noise of individual, or average individual, dynamics, so that +/// relevant variance is within-subject `asymDIFFUSION` +/// `p = −q / (2 a)`. The 2017-era `summary.ctsemFit.R` forms +/// unstandardised `DIFFUSION` as `mxEval(DIFFUSION, mxobj, +/// compute=TRUE)` and forms `asymDIFFUSIONstd` when +/// `verbose = TRUE`. That source does not form a `DIFFUSIONstd` +/// matrix; the scalar map here is the footnote 4 standardisation +/// of that named continuous diffusion: `q / p` after strictly +/// positive `p`. Form strictly positive `p` first, then divide +/// `q` by `p`. In the scalar stationary case that ratio equals +/// `−2 a`. Unstandardised `q` is defined for growing `a ≥ 0` and +/// for zero diffusion; standardised `DIFFUSION` is not. Zero `q` +/// has no positive process SD and fails closed. Lasting `p` +/// requires stable `a < 0`. Diffusion is an event-time +/// process-dynamics quantity, so a non-event clock fails closed. +/// `Q_Δt / p` is `discreteDIFFUSIONstd` and depends on the event +/// interval. `p / p = 1` is `asymDIFFUSIONstd` and recovers the +/// same number when `a = −0.5` and remains a distinct named +/// quantity. `q / (trait + p + added)` uses the total, not +/// `asymDIFFUSION`, and is not this map when `TRAITVAR` is +/// nonzero. This crate does not currently export +/// `discreteDIFFUSIONstd`; the refuse still names that quantity. +/// This is not a Kalman filter, not a matrix `expm`, not DSEM, +/// and not ctsem estimation. +/// +/// # Errors +/// +/// Returns [`PsychometricError::EventTimeRequired`] for any +/// non-event clock, +/// [`PsychometricError::StationaryVarianceRequiresStableDrift`] +/// when `a ≥ 0`, +/// [`PsychometricError::StandardisedContinuousDiffusionRequiresPositiveStationaryVariance`] +/// when `q = 0`, and +/// [`PsychometricError::InvalidNumericInput`] when the diffusion or +/// log-rate is non-finite, the diffusion is negative, or the +/// quotient overflows. +pub fn recover_standardised_continuous_diffusion( + continuous_diffusion: f64, + log_rate: f64, + clock: LagClock, +) -> Result { + let stationary = recover_stationary_latent_variance(continuous_diffusion, log_rate, clock)?; + if stationary == 0.0 { + return Err( + PsychometricError::StandardisedContinuousDiffusionRequiresPositiveStationaryVariance, + ); + } + require_finite(continuous_diffusion / stationary) +} + +/// Refuse treating unstandardised `DIFFUSION` as p. 16 +/// `DIFFUSIONstd`. +/// +/// Unstandardised `q` is defined for growing `a ≥ 0` and for a +/// zero process. Footnote 4 `DIFFUSIONstd` requires strictly +/// positive `asymDIFFUSION`. Equal numbers when `q = −2 a` are +/// still distinct named quantities. +/// +/// # Errors +/// +/// Always returns +/// [`PsychometricError::UnstandardisedContinuousDiffusionIsNotStandardisedContinuousDiffusion`]. +pub fn refuse_unstandardised_continuous_diffusion_as_standardised_continuous_diffusion( + unstandardised_diffusion: f64, + standardised_diffusion: f64, +) -> Result { + let _ = (unstandardised_diffusion, standardised_diffusion); + Err(PsychometricError::UnstandardisedContinuousDiffusionIsNotStandardisedContinuousDiffusion) +} + +/// Refuse treating p. 16 `discreteDIFFUSIONstd` as p. 16 +/// `DIFFUSIONstd`. +/// +/// `Q_Δt / p` depends on the event interval. `DIFFUSIONstd` +/// `q / p` does not. This crate does not currently export +/// `discreteDIFFUSIONstd`; the refuse still names that quantity. +/// +/// # Errors +/// +/// Always returns +/// [`PsychometricError::DiscreteStandardisedDiffusionIsNotStandardisedContinuousDiffusion`]. +pub fn refuse_discrete_standardised_diffusion_as_standardised_continuous_diffusion( + discrete_standardised_diffusion: f64, + standardised_continuous_diffusion: f64, +) -> Result { + let _ = ( + discrete_standardised_diffusion, + standardised_continuous_diffusion, + ); + Err(PsychometricError::DiscreteStandardisedDiffusionIsNotStandardisedContinuousDiffusion) +} + +/// Refuse treating `q / (trait + p + added)` as p. 16 +/// `DIFFUSIONstd`. +/// +/// Footnote 4 uses only within-subject `asymDIFFUSION`. The total +/// includes `TRAITVAR` and `addedTIPREDVAR`. Equal numbers when +/// those extras are zero remain distinct named quantities. +/// +/// # Errors +/// +/// Always returns +/// [`PsychometricError::TotalVarianceScaledDiffusionIsNotStandardisedContinuousDiffusion`]. +pub fn refuse_total_variance_scaled_diffusion_as_standardised_continuous_diffusion( + total_scaled_diffusion: f64, + standardised_continuous_diffusion: f64, +) -> Result { + let _ = (total_scaled_diffusion, standardised_continuous_diffusion); + Err(PsychometricError::TotalVarianceScaledDiffusionIsNotStandardisedContinuousDiffusion) +} + /// Refuse treating p. 16 `TIPREDVARstd` as p. 16 /// `asymDIFFUSIONstd`. /// @@ -6887,7 +7011,8 @@ mod tests { recover_manifest_lagged_observed_covariance, recover_manifest_observed_mean, recover_manifest_observed_variance, recover_manifest_trait_plus_state_observed_variance, recover_standardised_asymptotic_continuous_intercept, - recover_standardised_asymptotic_diffusion, recover_standardised_continuous_intercept, + recover_standardised_asymptotic_diffusion, recover_standardised_continuous_diffusion, + recover_standardised_continuous_intercept, recover_standardised_discrete_continuous_intercept, recover_standardised_initial_latent_mean, recover_standardised_initial_latent_variance, recover_standardised_manifest_mean, recover_standardised_manifest_trait_variance, @@ -6920,6 +7045,7 @@ mod tests { refuse_continuous_intercept_as_manifest_means, refuse_difference_quotient_as_local_rate, refuse_discrete_standardised_continuous_intercept_as_standardised_asymptotic_continuous_intercept, refuse_discrete_standardised_continuous_intercept_as_standardised_continuous_intercept, + refuse_discrete_standardised_diffusion_as_standardised_continuous_diffusion, refuse_evolved_observed_mean_as_after_extra_process_observed_mean, refuse_evolved_observed_mean_as_extra_process_observed_mean, refuse_evolved_observed_mean_as_impulse_carry_observed_mean, @@ -7026,12 +7152,14 @@ mod tests { refuse_time_independent_effect_as_time_varying_discrete_effect, refuse_time_independent_observed_mean_as_initial_time_dependent_observed_mean, refuse_time_independent_observed_mean_as_initial_time_independent_observed_mean, + refuse_total_variance_scaled_diffusion_as_standardised_continuous_diffusion, refuse_trait_plus_state_lagged_covariance_as_stationary_lagged_latent_covariance, refuse_trait_scaled_continuous_intercept_as_standardised_continuous_intercept, refuse_trait_variance_as_process_noise, refuse_trait_variance_as_stationary_within_subject, refuse_unmatched_time_varying_predictor_interval, refuse_unstandardised_asymptotic_continuous_intercept_as_standardised_asymptotic_continuous_intercept, refuse_unstandardised_asymptotic_diffusion_as_standardised_asymptotic_diffusion, + refuse_unstandardised_continuous_diffusion_as_standardised_continuous_diffusion, refuse_unstandardised_continuous_intercept_as_standardised_continuous_intercept, refuse_unstandardised_discrete_continuous_intercept_as_standardised_discrete_continuous_intercept, refuse_unstandardised_initial_latent_mean_as_standardised_initial_latent_mean, @@ -16033,6 +16161,124 @@ mod tests { ); } + #[test] + fn standardised_continuous_diffusion_recovers_driver_page_sixteen_ratio() { + // Driver et al. (2017, p. 16 DIFFUSIONstd; Table 2; footnote 4; + // Eq. 4; 2017-era summary.ctsemFit.R): form strictly positive + // asymDIFFUSION p = −q/(2a), then q/p. Scalar stationary map + // is −2a. That source does not form a DIFFUSIONstd matrix. + let diffusion = 0.4_f64; + let log_rate = -0.25_f64; + let recovered = + recover_standardised_continuous_diffusion(diffusion, log_rate, LagClock::EventTime) + .expect("DIFFUSIONstd"); + let stationary = + recover_stationary_latent_variance(diffusion, log_rate, LagClock::EventTime) + .expect("asymDIFFUSION"); + let expected = diffusion / stationary; + assert!((recovered - expected).abs() < 1e-15); + assert!((recovered - (-2.0 * log_rate)).abs() < 1e-15); + let larger_q = + recover_standardised_continuous_diffusion(1.6, log_rate, LagClock::EventTime) + .expect("DIFFUSIONstd q=1.6"); + assert!((larger_q - recovered).abs() < 1e-15); + assert!((diffusion - recovered).abs() > 1e-3); + let process_noise = + recover_discrete_process_noise(diffusion, log_rate, 1.0, LagClock::EventTime) + .expect("discreteDIFFUSION"); + let discrete_std = process_noise / stationary; + assert!((discrete_std - recovered).abs() > 1e-3); + let total = recover_trait_plus_state_latent_variance(0.5, stationary).expect("trait + p"); + let total_scaled = diffusion / total; + assert!((total_scaled - recovered).abs() > 1e-3); + let equal_ratio = recover_standardised_continuous_diffusion(0.4, -0.5, LagClock::EventTime) + .expect("a=-0.5"); + let equal_correlation = + recover_standardised_asymptotic_diffusion(0.4, -0.5, LagClock::EventTime) + .expect("asymDIFFUSIONstd a=-0.5"); + assert!((equal_ratio - equal_correlation).abs() < 1e-15); + let max_ratio = + recover_standardised_continuous_diffusion(f64::MAX, -0.75, LagClock::EventTime) + .expect("MAX/1.5"); + assert!((max_ratio - 1.5).abs() < 1e-15); + assert_eq!( + refuse_unstandardised_continuous_diffusion_as_standardised_continuous_diffusion( + diffusion, recovered + ), + Err( + PsychometricError::UnstandardisedContinuousDiffusionIsNotStandardisedContinuousDiffusion + ) + ); + assert_eq!( + refuse_discrete_standardised_diffusion_as_standardised_continuous_diffusion( + discrete_std, recovered + ), + Err( + PsychometricError::DiscreteStandardisedDiffusionIsNotStandardisedContinuousDiffusion + ) + ); + assert_eq!( + refuse_total_variance_scaled_diffusion_as_standardised_continuous_diffusion( + total_scaled, + recovered + ), + Err( + PsychometricError::TotalVarianceScaledDiffusionIsNotStandardisedContinuousDiffusion + ) + ); + assert_eq!( + refuse_standardised_continuous_diffusion_as_standardised_asymptotic_diffusion( + equal_ratio, + equal_correlation + ), + Err( + PsychometricError::StandardisedContinuousDiffusionIsNotStandardisedAsymptoticDiffusion + ) + ); + } + + #[test] + fn standardised_continuous_diffusion_fails_closed_when_unstandardised_is_defined() { + assert_eq!( + recover_standardised_continuous_diffusion(0.0, -0.25, LagClock::EventTime), + Err( + PsychometricError::StandardisedContinuousDiffusionRequiresPositiveStationaryVariance + ) + ); + assert_eq!( + recover_standardised_continuous_diffusion(0.4, -0.25, LagClock::SystemTime), + Err(PsychometricError::EventTimeRequired) + ); + assert_eq!( + recover_standardised_continuous_diffusion(0.4, 0.25, LagClock::EventTime), + Err(PsychometricError::StationaryVarianceRequiresStableDrift) + ); + assert_eq!( + recover_standardised_continuous_diffusion(0.4, 0.0, LagClock::EventTime), + Err(PsychometricError::StationaryVarianceRequiresStableDrift) + ); + assert_eq!( + recover_standardised_continuous_diffusion(-0.4, -0.25, LagClock::EventTime), + Err(PsychometricError::InvalidNumericInput) + ); + assert_eq!( + recover_standardised_continuous_diffusion(f64::NAN, -0.25, LagClock::EventTime), + Err(PsychometricError::InvalidNumericInput) + ); + assert_eq!( + recover_standardised_continuous_diffusion(0.4, f64::NAN, LagClock::EventTime), + Err(PsychometricError::InvalidNumericInput) + ); + assert_eq!( + recover_standardised_continuous_diffusion(f64::INFINITY, -0.25, LagClock::EventTime), + Err(PsychometricError::InvalidNumericInput) + ); + assert_eq!( + recover_standardised_continuous_diffusion(1.0, -1e308, LagClock::EventTime), + Err(PsychometricError::InvalidNumericInput) + ); + } + #[test] fn standardised_manifest_trait_variance_recovers_driver_table_two_after_positive_psi() { // Driver et al. (2017, Table 2 MANIFESTTRAITVAR; §7.1; p. 16 diff --git a/crates/psychometric_core/src/lib.rs b/crates/psychometric_core/src/lib.rs index c081a63f6..a1e2f70d8 100644 --- a/crates/psychometric_core/src/lib.rs +++ b/crates/psychometric_core/src/lib.rs @@ -246,6 +246,17 @@ //! is `TIPREDVARstd` and is not that map even when both equal 1; //! zero `q` fails closed; a non-event clock fails closed; `a ≥ 0` //! fails closed; JSS PDF re-opened 2026-08-26T17:20Z), +//! recovers the Driver p. 16 `DIFFUSIONstd` as `q / p = −2 a` after +//! strictly positive `asymDIFFUSION` `p = −q / (2 a)` (footnote 4 +//! uses only the relevant within-subject variance; 2017-era +//! `summary.ctsemFit.R` forms unstandardised `DIFFUSION` and +//! `asymDIFFUSIONstd` and does not form a `DIFFUSIONstd` matrix; +//! unstandardised `q` is defined for growing `a ≥ 0` and for a zero +//! process and is not that map; `Q_Δt / p` is +//! `discreteDIFFUSIONstd` and is not that continuous ratio; +//! `q / (trait + p + added)` uses the total and is not that map; +//! zero `q` fails closed; a non-event clock fails closed; `a ≥ 0` +//! fails closed; JSS PDF re-opened 2026-09-01T19:20Z), //! recovers the Driver p. 16 `MANIFESTTRAITVARstd` as `ψ / ψ = 1` //! after strictly positive `MANIFESTTRAITVAR` (Table 2 names //! `MANIFESTTRAITVAR` `Ψ_τ`; §7.1 names manifest traits stable @@ -403,6 +414,8 @@ pub use event_time::recover_manifest_trait_plus_state_observed_variance; pub use event_time::recover_standardised_asymptotic_continuous_intercept; /// Exact scalar p. 16 `asymDIFFUSIONstd` `p / p = 1` after strictly positive `asymDIFFUSION`. pub use event_time::recover_standardised_asymptotic_diffusion; +/// Exact scalar p. 16 `DIFFUSIONstd` `q / p = −2 a` after strictly positive `asymDIFFUSION`. +pub use event_time::recover_standardised_continuous_diffusion; /// Exact scalar p. 16 `CINTstd` `κ / √p`. pub use event_time::recover_standardised_continuous_intercept; /// Exact scalar p. 16 `discreteCINTstd` `A^{-1}[e^{A Δt} − I] κ / √p`. @@ -492,6 +505,8 @@ pub use event_time::refuse_difference_quotient_as_local_rate; pub use event_time::refuse_discrete_standardised_continuous_intercept_as_standardised_asymptotic_continuous_intercept; /// Refuse treating p. 16 `discreteCINTstd` as `CINTstd`. pub use event_time::refuse_discrete_standardised_continuous_intercept_as_standardised_continuous_intercept; +/// Refuse treating p. 16 `discreteDIFFUSIONstd` as `DIFFUSIONstd`. +pub use event_time::refuse_discrete_standardised_diffusion_as_standardised_continuous_diffusion; /// Refuse treating evolved `τ + λ μ_t` as the after-t0 extra-process observed mean. pub use event_time::refuse_evolved_observed_mean_as_after_extra_process_observed_mean; /// Refuse treating evolved `τ + λ μ_t` as the extra-process observed mean. @@ -636,6 +651,10 @@ pub use event_time::refuse_standardised_initial_latent_variance_as_standardised_ pub use event_time::refuse_standardised_manifest_trait_variance_as_standardised_manifest_variance; /// Refuse treating `MANIFESTVARstd` as `MANIFESTMEANSstd`. pub use event_time::refuse_standardised_manifest_variance_as_standardised_manifest_mean; +/// Refuse treating total-variance scaled `q / (trait + p + added)` as `DIFFUSIONstd`. +pub use event_time::refuse_total_variance_scaled_diffusion_as_standardised_continuous_diffusion; +/// Refuse treating unstandardised `DIFFUSION` as p. 16 `DIFFUSIONstd`. +pub use event_time::refuse_unstandardised_continuous_diffusion_as_standardised_continuous_diffusion; /// Refuse treating observed θ as p. 16 `MANIFESTVARstd`. pub use event_time::refuse_observed_variance_as_standardised_manifest_variance; diff --git a/crates/psychometric_core/tests/multilevel_event_time_recovery_contract.rs b/crates/psychometric_core/tests/multilevel_event_time_recovery_contract.rs index 1c0027f44..7aae42eb0 100644 --- a/crates/psychometric_core/tests/multilevel_event_time_recovery_contract.rs +++ b/crates/psychometric_core/tests/multilevel_event_time_recovery_contract.rs @@ -35,18 +35,18 @@ use psychometric_core::{ recover_manifest_lagged_observed_covariance, recover_manifest_observed_mean, recover_manifest_observed_variance, recover_manifest_trait_plus_state_observed_variance, recover_standardised_asymptotic_continuous_intercept, - recover_standardised_asymptotic_diffusion, recover_standardised_continuous_intercept, - recover_standardised_discrete_continuous_intercept, recover_standardised_initial_latent_mean, - recover_standardised_initial_latent_variance, recover_standardised_manifest_mean, - recover_standardised_manifest_trait_variance, recover_standardised_manifest_variance, - recover_standardised_trait_variance, recover_stationary_initial_latent_mean, - recover_stationary_initial_latent_variance, recover_stationary_initial_observed_mean, - recover_stationary_initial_observed_variance, recover_stationary_lagged_latent_covariance, - recover_stationary_lagged_observed_covariance, recover_stationary_latent_variance, - recover_stationary_later_latent_variance, recover_stationary_later_observed_variance, - recover_time_dependent_predictor_impulse, recover_time_dependent_predictor_impulse_carry, - recover_trait_plus_state_lagged_covariance, recover_trait_plus_state_latent_variance, - recover_within_residual_event_time_log_rate, + recover_standardised_asymptotic_diffusion, recover_standardised_continuous_diffusion, + recover_standardised_continuous_intercept, recover_standardised_discrete_continuous_intercept, + recover_standardised_initial_latent_mean, recover_standardised_initial_latent_variance, + recover_standardised_manifest_mean, recover_standardised_manifest_trait_variance, + recover_standardised_manifest_variance, recover_standardised_trait_variance, + recover_stationary_initial_latent_mean, recover_stationary_initial_latent_variance, + recover_stationary_initial_observed_mean, recover_stationary_initial_observed_variance, + recover_stationary_lagged_latent_covariance, recover_stationary_lagged_observed_covariance, + recover_stationary_latent_variance, recover_stationary_later_latent_variance, + recover_stationary_later_observed_variance, recover_time_dependent_predictor_impulse, + recover_time_dependent_predictor_impulse_carry, recover_trait_plus_state_lagged_covariance, + recover_trait_plus_state_latent_variance, recover_within_residual_event_time_log_rate, refuse_after_extra_process_contribution_as_observed_mean, refuse_after_extra_process_latent_mean_as_observed_mean, refuse_asymptotic_continuous_intercept_as_asymptotic_time_independent_effect, @@ -64,6 +64,7 @@ use psychometric_core::{ refuse_continuous_intercept_as_discrete_mean_increment, refuse_continuous_intercept_as_initial_latent_mean, refuse_continuous_intercept_as_manifest_means, refuse_difference_quotient_as_local_rate, + refuse_discrete_standardised_diffusion_as_standardised_continuous_diffusion, refuse_evolved_observed_mean_as_after_extra_process_observed_mean, refuse_evolved_observed_mean_as_extra_process_observed_mean, refuse_evolved_observed_mean_as_impulse_carry_observed_mean, @@ -159,9 +160,11 @@ use psychometric_core::{ refuse_time_independent_effect_as_time_varying_discrete_effect, refuse_time_independent_observed_mean_as_initial_time_dependent_observed_mean, refuse_time_independent_observed_mean_as_initial_time_independent_observed_mean, + refuse_total_variance_scaled_diffusion_as_standardised_continuous_diffusion, refuse_trait_plus_state_lagged_covariance_as_stationary_lagged_latent_covariance, refuse_trait_variance_as_process_noise, refuse_trait_variance_as_stationary_within_subject, refuse_unmatched_time_varying_predictor_interval, + refuse_unstandardised_continuous_diffusion_as_standardised_continuous_diffusion, refuse_unstandardised_manifest_trait_variance_as_standardised_manifest_trait_variance, refuse_unstandardised_manifest_variance_as_standardised_manifest_variance, refuse_unstandardised_trait_variance_as_standardised_trait_variance, @@ -6158,6 +6161,100 @@ fn standardised_asymptotic_diffusion_refuses_non_event_clocks_and_does_not_keep_ ); } +#[test] +fn standardised_continuous_diffusion_recovers_driver_page_sixteen_ratio() { + let diffusion = 0.4_f64; + let log_rate = -0.25_f64; + let recovered = + recover_standardised_continuous_diffusion(diffusion, log_rate, LagClock::EventTime) + .expect("DIFFUSIONstd"); + let recovered_error = (recovered - (-2.0 * log_rate)).abs(); + assert!( + recovered_error < 1e-15, + "Driver et al. (2017, p. 16 DIFFUSIONstd): RMSE {recovered_error} for q / p = −2 a" + ); + let larger_q = recover_standardised_continuous_diffusion(1.6, log_rate, LagClock::EventTime) + .expect("DIFFUSIONstd q=1.6"); + assert!( + (larger_q - recovered).abs() < 1e-15, + "Driver et al. (2017, p. 16): distinct positive q recover the same DIFFUSIONstd" + ); + let unstandardised_error = (diffusion - (-2.0 * log_rate)).abs(); + assert!( + unstandardised_error > recovered_error, + "Driver et al. (2017, Table 2): unstandardised DIFFUSION RMSE {unstandardised_error} must exceed DIFFUSIONstd RMSE {recovered_error}" + ); + let stationary = + recover_stationary_latent_variance(diffusion, log_rate, LagClock::EventTime).expect("p"); + let process_noise = + recover_discrete_process_noise(diffusion, log_rate, 1.0, LagClock::EventTime) + .expect("discreteDIFFUSION"); + let discrete_std = process_noise / stationary; + let discrete_error = (discrete_std - (-2.0 * log_rate)).abs(); + assert!( + discrete_error > recovered_error, + "Driver et al. (2017, p. 16): discreteDIFFUSIONstd RMSE {discrete_error} must exceed DIFFUSIONstd RMSE {recovered_error}" + ); + let total = recover_trait_plus_state_latent_variance(0.5, stationary).expect("trait + p"); + let total_scaled = diffusion / total; + let total_error = (total_scaled - (-2.0 * log_rate)).abs(); + assert!( + total_error > recovered_error, + "Driver et al. (2017, footnote 4): total-scaled RMSE {total_error} must exceed DIFFUSIONstd RMSE {recovered_error}" + ); + let equal_ratio = + recover_standardised_continuous_diffusion(0.4, -0.5, LagClock::EventTime).expect("a=-0.5"); + let equal_correlation = + recover_standardised_asymptotic_diffusion(0.4, -0.5, LagClock::EventTime) + .expect("asymDIFFUSIONstd a=-0.5"); + assert!( + (equal_ratio - equal_correlation).abs() < 1e-15, + "Driver et al. (2017, p. 16): a = −0.5 makes DIFFUSIONstd equal 1 with asymDIFFUSIONstd" + ); + assert_eq!( + refuse_unstandardised_continuous_diffusion_as_standardised_continuous_diffusion( + diffusion, recovered + ), + Err( + PsychometricError::UnstandardisedContinuousDiffusionIsNotStandardisedContinuousDiffusion + ) + ); + assert_eq!( + refuse_discrete_standardised_diffusion_as_standardised_continuous_diffusion( + discrete_std, + recovered + ), + Err(PsychometricError::DiscreteStandardisedDiffusionIsNotStandardisedContinuousDiffusion) + ); + assert_eq!( + refuse_total_variance_scaled_diffusion_as_standardised_continuous_diffusion( + total_scaled, + recovered + ), + Err(PsychometricError::TotalVarianceScaledDiffusionIsNotStandardisedContinuousDiffusion) + ); +} + +#[test] +fn standardised_continuous_diffusion_refuses_non_event_clocks_and_does_not_keep_zero_q() { + assert_eq!( + recover_standardised_continuous_diffusion(0.4, -0.25, LagClock::AssertionTime), + Err(PsychometricError::EventTimeRequired) + ); + assert_eq!( + recover_standardised_continuous_diffusion(0.4, -0.25, LagClock::KnowledgeCutoff), + Err(PsychometricError::EventTimeRequired) + ); + assert_eq!( + recover_standardised_continuous_diffusion(0.0, -0.25, LagClock::EventTime), + Err(PsychometricError::StandardisedContinuousDiffusionRequiresPositiveStationaryVariance) + ); + assert_eq!( + recover_standardised_continuous_diffusion(0.4, 0.25, LagClock::EventTime), + Err(PsychometricError::StationaryVarianceRequiresStableDrift) + ); +} + #[test] fn standardised_manifest_trait_variance_recovers_driver_table_two_correlation() { let manifest_trait = 1.6_f64; diff --git a/crates/psychometric_core/tests/scientific_claim_boundary_contract.rs b/crates/psychometric_core/tests/scientific_claim_boundary_contract.rs index 6ccf7f38b..563732968 100644 --- a/crates/psychometric_core/tests/scientific_claim_boundary_contract.rs +++ b/crates/psychometric_core/tests/scientific_claim_boundary_contract.rs @@ -30,18 +30,18 @@ use psychometric_core::{ recover_manifest_lagged_observed_covariance, recover_manifest_observed_mean, recover_manifest_observed_variance, recover_manifest_trait_plus_state_observed_variance, recover_standardised_asymptotic_continuous_intercept, - recover_standardised_asymptotic_diffusion, recover_standardised_continuous_intercept, - recover_standardised_discrete_continuous_intercept, recover_standardised_initial_latent_mean, - recover_standardised_initial_latent_variance, recover_standardised_manifest_mean, - recover_standardised_manifest_trait_variance, recover_standardised_manifest_variance, - recover_standardised_trait_variance, recover_stationary_initial_latent_mean, - recover_stationary_initial_latent_variance, recover_stationary_initial_observed_mean, - recover_stationary_initial_observed_variance, recover_stationary_lagged_latent_covariance, - recover_stationary_lagged_observed_covariance, recover_stationary_latent_variance, - recover_stationary_later_latent_variance, recover_stationary_later_observed_variance, - recover_time_dependent_predictor_impulse, recover_time_dependent_predictor_impulse_carry, - recover_trait_plus_state_lagged_covariance, recover_trait_plus_state_latent_variance, - recover_within_residual_event_time_log_rate, + recover_standardised_asymptotic_diffusion, recover_standardised_continuous_diffusion, + recover_standardised_continuous_intercept, recover_standardised_discrete_continuous_intercept, + recover_standardised_initial_latent_mean, recover_standardised_initial_latent_variance, + recover_standardised_manifest_mean, recover_standardised_manifest_trait_variance, + recover_standardised_manifest_variance, recover_standardised_trait_variance, + recover_stationary_initial_latent_mean, recover_stationary_initial_latent_variance, + recover_stationary_initial_observed_mean, recover_stationary_initial_observed_variance, + recover_stationary_lagged_latent_covariance, recover_stationary_lagged_observed_covariance, + recover_stationary_latent_variance, recover_stationary_later_latent_variance, + recover_stationary_later_observed_variance, recover_time_dependent_predictor_impulse, + recover_time_dependent_predictor_impulse_carry, recover_trait_plus_state_lagged_covariance, + recover_trait_plus_state_latent_variance, recover_within_residual_event_time_log_rate, refuse_after_extra_process_contribution_as_observed_mean, refuse_after_extra_process_latent_mean_as_observed_mean, refuse_asymptotic_continuous_intercept_as_asymptotic_time_independent_effect, @@ -63,6 +63,7 @@ use psychometric_core::{ refuse_continuous_intercept_as_manifest_means, refuse_discrete_standardised_continuous_intercept_as_standardised_asymptotic_continuous_intercept, refuse_discrete_standardised_continuous_intercept_as_standardised_continuous_intercept, + refuse_discrete_standardised_diffusion_as_standardised_continuous_diffusion, refuse_evolved_observed_mean_as_after_extra_process_observed_mean, refuse_evolved_observed_mean_as_extra_process_observed_mean, refuse_evolved_observed_mean_as_impulse_carry_observed_mean, @@ -167,11 +168,13 @@ use psychometric_core::{ refuse_time_independent_effect_as_time_varying_discrete_effect, refuse_time_independent_observed_mean_as_initial_time_dependent_observed_mean, refuse_time_independent_observed_mean_as_initial_time_independent_observed_mean, + refuse_total_variance_scaled_diffusion_as_standardised_continuous_diffusion, refuse_trait_plus_state_lagged_covariance_as_stationary_lagged_latent_covariance, refuse_trait_scaled_continuous_intercept_as_standardised_continuous_intercept, refuse_trait_variance_as_process_noise, refuse_trait_variance_as_stationary_within_subject, refuse_unstandardised_asymptotic_continuous_intercept_as_standardised_asymptotic_continuous_intercept, refuse_unstandardised_asymptotic_diffusion_as_standardised_asymptotic_diffusion, + refuse_unstandardised_continuous_diffusion_as_standardised_continuous_diffusion, refuse_unstandardised_continuous_intercept_as_standardised_continuous_intercept, refuse_unstandardised_discrete_continuous_intercept_as_standardised_discrete_continuous_intercept, refuse_unstandardised_initial_latent_mean_as_standardised_initial_latent_mean, @@ -3389,6 +3392,101 @@ fn standardised_asymptotic_diffusion_is_not_unstandardised_t0var_or_diffusion_ra ); } +#[test] +fn standardised_continuous_diffusion_is_not_unstandardised_discrete_or_total() { + let diffusion = 0.4_f64; + let log_rate = -0.25_f64; + let recovered = + recover_standardised_continuous_diffusion(diffusion, log_rate, LagClock::EventTime) + .expect("DIFFUSIONstd"); + assert!( + (recovered - (-2.0 * log_rate)).abs() < 1e-15, + "Driver et al. (2017, p. 16 footnote 4): DIFFUSIONstd is q/p = −2a" + ); + let larger_q = recover_standardised_continuous_diffusion(1.6, log_rate, LagClock::EventTime) + .expect("DIFFUSIONstd q=1.6"); + assert!( + (larger_q - recovered).abs() < 1e-15, + "Driver et al. (2017, p. 16): distinct positive q recover the same DIFFUSIONstd" + ); + assert!( + (recovered - diffusion).abs() > 1e-3, + "Driver et al. (2017, Table 2): unstandardised DIFFUSION is not DIFFUSIONstd" + ); + let stationary = + recover_stationary_latent_variance(diffusion, log_rate, LagClock::EventTime).expect("p"); + let process_noise = + recover_discrete_process_noise(diffusion, log_rate, 1.0, LagClock::EventTime) + .expect("discreteDIFFUSION"); + let discrete_std = process_noise / stationary; + assert!( + (discrete_std - recovered).abs() > 1e-3, + "Driver et al. (2017, p. 16): discreteDIFFUSIONstd is not DIFFUSIONstd" + ); + let total = recover_trait_plus_state_latent_variance(0.5, stationary).expect("trait + p"); + let total_scaled = diffusion / total; + assert!( + (total_scaled - recovered).abs() > 1e-3, + "Driver et al. (2017, footnote 4): q / (trait + p + added) is not DIFFUSIONstd" + ); + let equal_ratio = + recover_standardised_continuous_diffusion(0.4, -0.5, LagClock::EventTime).expect("a=-0.5"); + let equal_correlation = + recover_standardised_asymptotic_diffusion(0.4, -0.5, LagClock::EventTime) + .expect("asymDIFFUSIONstd a=-0.5"); + assert!( + (equal_ratio - equal_correlation).abs() < 1e-15, + "Driver et al. (2017, p. 16): equal 1 when a = −0.5 remains a distinct named quantity" + ); + assert_eq!( + refuse_unstandardised_continuous_diffusion_as_standardised_continuous_diffusion( + diffusion, recovered + ), + Err( + psychometric_core::PsychometricError::UnstandardisedContinuousDiffusionIsNotStandardisedContinuousDiffusion + ) + ); + assert_eq!( + refuse_discrete_standardised_diffusion_as_standardised_continuous_diffusion( + discrete_std, recovered + ), + Err( + psychometric_core::PsychometricError::DiscreteStandardisedDiffusionIsNotStandardisedContinuousDiffusion + ) + ); + assert_eq!( + refuse_total_variance_scaled_diffusion_as_standardised_continuous_diffusion( + total_scaled, recovered + ), + Err( + psychometric_core::PsychometricError::TotalVarianceScaledDiffusionIsNotStandardisedContinuousDiffusion + ) + ); + assert_eq!( + refuse_standardised_continuous_diffusion_as_standardised_asymptotic_diffusion( + equal_ratio, + equal_correlation + ), + Err( + psychometric_core::PsychometricError::StandardisedContinuousDiffusionIsNotStandardisedAsymptoticDiffusion + ) + ); + assert_eq!( + recover_standardised_continuous_diffusion(0.0, log_rate, LagClock::EventTime), + Err( + psychometric_core::PsychometricError::StandardisedContinuousDiffusionRequiresPositiveStationaryVariance + ) + ); + assert_eq!( + recover_standardised_continuous_diffusion(diffusion, log_rate, LagClock::DocumentTime), + Err(psychometric_core::PsychometricError::EventTimeRequired) + ); + assert_eq!( + recover_standardised_continuous_diffusion(diffusion, 0.25, LagClock::EventTime), + Err(psychometric_core::PsychometricError::StationaryVarianceRequiresStableDrift) + ); +} + #[test] fn standardised_manifest_trait_variance_is_not_unstandardised_traitstd_or_measurement_error() { let manifest_trait = 1.6_f64; diff --git a/docs/adr/0005-posterior-esem-dsem.md b/docs/adr/0005-posterior-esem-dsem.md index ee1e6cf0d..870f35111 100644 --- a/docs/adr/0005-posterior-esem-dsem.md +++ b/docs/adr/0005-posterior-esem-dsem.md @@ -35,6 +35,7 @@ The executable standardised-asymptotic-intercept slice recovers Driver et al. (2 The executable standardised-initial-mean slice recovers Driver et al. (2017, p. 16 `T0MEANSstd`) as `μ_0 / √p_0` after strictly positive free `T0VAR` (footnote 4; JSS PDF re-opened 2026-08-26T04:09Z). Unstandardised `μ_0` is defined for a zero first-occasion variance and is not that map. `p_0 / p_0 = 1` is the named `T0VARstd` correlation form and is not `T0MEANSstd` even when `μ_0 = √p_0`. `μ_0 / √asymDIFFUSION` uses process-dynamics variance and is not the first-occasion map. Free `T0MEANS` does not require `a < 0`. This is not ctsem estimation. The executable standardised-initial-variance slice recovers Driver et al. (2017, p. 16 `T0VARstd`) as `p_0 / p_0 = 1` after strictly positive free `T0VAR` (footnote 4; 2017-era `summary.ctsemFit.R` `solve(sqrt(diag(T0VAR))) %&% T0VAR`; JSS PDF re-opened 2026-08-26T07:17Z). Unstandardised `p_0` is defined for a zero first-occasion variance and is not that map. `μ_0 / √p_0` is the named `T0MEANSstd` first-occasion mean and is not `T0VARstd` even when `μ_0 = √p_0`. `p / p = 1` is the named `asymDIFFUSIONstd` correlation form and is not `T0VARstd` even when both equal 1. Free `T0VAR` does not require `a < 0`. This is not ctsem estimation. The executable standardised-asymptotic-diffusion slice recovers Driver et al. (2017, p. 16 `asymDIFFUSIONstd`) as `p / p = 1` after strictly positive `asymDIFFUSION` `p = −q / (2 a)` (footnote 4; 2017-era `summary.ctsemFit.R` `solve(sqrt(diag(asymDIFFUSION))) %&% asymDIFFUSION`; JSS PDF re-opened 2026-08-26T17:20Z). Unstandardised `p` is defined for a zero process and is not that map. `p_0 / p_0 = 1` is the named `T0VARstd` first-occasion correlation and is not `asymDIFFUSIONstd` even when both equal 1. `q / p = −2 a` is the named `DIFFUSIONstd` continuous-diffusion ratio and is not this correlation. `v / v = 1` is the named `TIPREDVARstd` predictor correlation and is not this map even when both equal 1. Zero `q` and `a ≥ 0` fail closed. This is not ctsem estimation. +The executable standardised-continuous-diffusion slice recovers Driver et al. (2017, p. 16 `DIFFUSIONstd`) as `q / p = −2 a` after strictly positive `asymDIFFUSION` `p = −q / (2 a)` (footnote 4; 2017-era `summary.ctsemFit.R` forms unstandardised `DIFFUSION` and `asymDIFFUSIONstd` and does not form a `DIFFUSIONstd` matrix; JSS PDF re-opened 2026-09-01T19:20Z). Unstandardised `q` is defined for growing `a ≥ 0` and for a zero process and is not that map. `Q_Δt / p` is `discreteDIFFUSIONstd` and is not this continuous ratio. `q / (trait + p + added)` uses the total and is not this map. `p / p = 1` is `asymDIFFUSIONstd` and is not this ratio even when both equal 1 (`a = −0.5`). Zero `q` and `a ≥ 0` fail closed. This is not ctsem estimation. The executable standardised-manifest-trait-variance slice recovers Driver et al. (2017, p. 16 `MANIFESTTRAITVARstd`) as `ψ / ψ = 1` after strictly positive `MANIFESTTRAITVAR` (Table 2 `Ψ_τ`; §7.1, p. 19; footnote 4; 2017-era `summary.ctsemFit.R` `solve(sqrt(diag(MANIFESTTRAITVAR) + ridging)) %&% MANIFESTTRAITVAR`; JSS PDF re-opened 2026-08-27T14:20Z). Unstandardised `ψ` is defined for a zero manifest trait and is not that map. `trait / trait = 1` is the named `TRAITVARstd` process-level correlation and is not `MANIFESTTRAITVARstd` even when both equal 1. `θ` is `MANIFESTVAR` and is not that correlation. `MANIFESTTRAITVAR` does not require `a < 0`. This is not ctsem estimation. The executable standardised-manifest-variance slice recovers Driver et al. (2017, p. 16 `MANIFESTVARstd`) as `θ / θ = 1` after strictly positive `MANIFESTVAR` (Table 2 measurement-error Cholesky; Eq. 5 `ε ~ N(0, Θ)`; footnote 4; 2017-era `summary.ctsemFit.R` `solve(sqrt(diag(MANIFESTVAR) + ridging)) %&% MANIFESTVAR`; default ridge 0; JSS PDF re-opened 2026-08-27T14:25Z). Unstandardised `θ` is defined for a zero residual and is not that map. `ψ / ψ = 1` is `MANIFESTTRAITVARstd` and is not `MANIFESTVARstd` even when both equal 1. `MANIFESTVAR` does not require `a < 0`. diff --git a/docs/research/multilevel-event-time-recovery.md b/docs/research/multilevel-event-time-recovery.md index 3701dcb4b..6877d1c62 100644 --- a/docs/research/multilevel-event-time-recovery.md +++ b/docs/research/multilevel-event-time-recovery.md @@ -62,7 +62,7 @@ This slice stays inside `psychometric_core`. It does not add a second invariance 56. refuse treating unstandardised `discreteDRIFT` `e^{a Δt}` as `discreteDRIFTstd`, refuse treating the §7.1 trait-plus-state autocorrelation `(trait + e^{a Δt} p + added) / (trait + p + added)` as `discreteDRIFTstd`, and refuse treating `TRAITVAR` as the footnote 4 standardisation variance; 57. recover the exact scalar p. 16 `discreteDIFFUSIONstd` `Q_Δt / (−q / (2 a))` after forming strictly positive `asymDIFFUSION` `−q / (2 a)` (Driver et al., 2017, p. 16; Eq. 3–4, pp. 4–5; footnote 4; §7.1, pp. 18–19; JSS PDF re-opened 2026-08-23T13:06Z; form the within-subject variance first, then `Q_Δt`, then the ratio; scalar stationary map is `1 − exp(2 a Δt)`; `a ≥ 0` and `q = 0` fail closed); 58. refuse treating unstandardised `discreteDIFFUSION` `Q_Δt` as `discreteDIFFUSIONstd`, refuse treating the continuous standardisation `q / (−q / (2 a)) = −2 a` as `discreteDIFFUSIONstd`, refuse treating `Q_Δt / (trait + p + added)` as `discreteDIFFUSIONstd`, and refuse treating `TRAITVAR` as the footnote 4 standardisation variance; -59. recover the exact scalar p. 16 `DIFFUSIONstd` `q / (−q / (2 a)) = −2 a` after forming strictly positive `asymDIFFUSION` `−q / (2 a)` (Driver et al., 2017, p. 16; Eq. 4, p. 5; footnote 4; §7.1, pp. 18–19; JSS PDF re-opened 2026-08-23T13:20Z; form the within-subject variance first, then `q / p`; scalar stationary map is `−2 a` and does not depend on `q` once `q > 0`; `a ≥ 0` and `q = 0` fail closed); +59. recover the exact scalar p. 16 `DIFFUSIONstd` `q / (−q / (2 a)) = −2 a` after forming strictly positive `asymDIFFUSION` `−q / (2 a)` (Driver et al., 2017, p. 16; Eq. 4, p. 5; Table 2, p. 12; footnote 4; §7.1, pp. 18–19; 2017-era ctsem `summary.ctsemFit.R`; JSS PDF re-opened 2026-09-01T19:20Z; that source forms unstandardised `DIFFUSION` and `asymDIFFUSIONstd` and does not form a `DIFFUSIONstd` matrix; form the within-subject variance first, then `q / p`; scalar stationary map is `−2 a` and does not depend on `q` once `q > 0`; `a ≥ 0` and `q = 0` fail closed); 60. refuse treating unstandardised `DIFFUSION` `q` as `DIFFUSIONstd`, refuse treating the discrete standardisation `Q_Δt / (−q / (2 a)) = 1 − exp(2 a Δt)` as `DIFFUSIONstd`, refuse treating `q / (trait + p + added)` as `DIFFUSIONstd`, and refuse treating `TRAITVAR` as the footnote 4 standardisation variance; 61. recover the exact scalar p. 16 `DRIFTstd` after forming strictly positive `asymDIFFUSION` `−q / (2 a)` (Driver et al., 2017, p. 16; Eq. 1, p. 4; footnote 4; §7.1, pp. 18–19; JSS PDF re-opened 2026-08-23T13:28Z; form the within-subject variance first; scalar stationary SD ratio is 1 so the standardised auto-effect equals `a` numerically; those remain distinct named quantities; `a ≥ 0` and `q = 0` fail closed); 62. refuse treating unstandardised `DRIFT` `a` as `DRIFTstd`, refuse treating the discrete standardisation `e^{a Δt}` as `DRIFTstd`, refuse treating `a p / (trait + p + added)` as `DRIFTstd`, and refuse treating `TRAITVAR` as the footnote 4 standardisation variance; @@ -240,7 +240,7 @@ The Voelkle et al. (2012) ZORA accepted manuscript was re-opened 2026-08-18T21:0 - already-centered irregular residuals recover a known drift at machine-scale RMSE, and that RMSE is smaller than CWC of the corresponding raw autoregressive series (Curran & Bauer, 2011, pp. 607–608); - a singleton cluster is skipped; two singleton clusters yield an empty pair list and fail closed; - overflowing CWC residuals, overflowing contextual subtraction, later-only residual overflow, non-finite intervals, Newton overflow / start-skip / deriv-INF, and Pearson empty/mismatch paths fail closed. -- Driver et al. (2017, p. 16 `DIFFUSIONstd`; Eq. 4; footnote 4; JSS PDF re-opened 2026-08-23T13:20Z) recovers a known continuous standardisation \(q/(-q/(2a))=-2a\) at machine-scale RMSE, and that RMSE is smaller than treating unstandardised \(q\), discrete \(Q_{\Delta t}/p\), or \(q/(\mathrm{trait}+p+\mathrm{added})\) as `DIFFUSIONstd`; distinct positive \(q\) recover the same \(-2a\); \(q=0\) and \(a\ge 0\) fail closed; a non-event clock and an overflowing ratio fail closed. +- Driver et al. (2017, p. 16 `DIFFUSIONstd`; Eq. 4; Table 2; footnote 4; 2017-era ctsem `summary.ctsemFit.R`; JSS PDF re-opened 2026-09-01T19:20Z) recovers a known continuous standardisation \(q/(-q/(2a))=-2a\) at machine-scale RMSE after strictly positive `asymDIFFUSION`, and that RMSE is smaller than treating unstandardised \(q\), discrete \(Q_{\Delta t}/p\), or \(q/(\mathrm{trait}+p+\mathrm{added})\) as `DIFFUSIONstd`; that 2017-era source does not form a `DIFFUSIONstd` matrix; distinct positive \(q\) recover the same \(-2a\); equal 1 with `asymDIFFUSIONstd` when \(a=-0.5\) remains a distinct named quantity; \(q=0\) and \(a\ge 0\) fail closed; a non-event clock and an overflowing ratio fail closed. - Driver et al. (2017, p. 16 `DRIFTstd`; Eq. 1; footnote 4; JSS PDF re-opened 2026-08-23T13:28Z) recovers a known continuous auto-effect \(a\) after strictly positive `asymDIFFUSION` at machine-scale RMSE, and that RMSE is smaller than treating unstandardised \(a\), discrete \(e^{a\Delta t}\), or \(ap/(\mathrm{trait}+p+\mathrm{added})\) as `DRIFTstd`; distinct positive \(q\) recover the same \(a\); \(q=0\) and \(a\ge 0\) fail closed; a non-event clock fails closed. - Driver et al. (2017, p. 16 `asymTIPREDEFFECTstd`; §7.2; Eq. 3; footnote 4; JSS PDF re-opened 2026-08-23T14:25Z) recovers a known standardised asymptotic TI effect \((-B/a)\cdot\sqrt{v}/\sqrt{-q/(2a)}\) at machine-scale RMSE, and that RMSE is smaller than treating unstandardised \(-B/a\), finite-interval \(A^{-1}[e^{A\Delta t}-I]B\cdot\sqrt{v}/\sqrt{p}\), or \((-B/a)\cdot\sqrt{v}/\sqrt{\mathrm{trait}+p+\mathrm{added}}\) as `asymTIPREDEFFECTstd`; a larger positive \(q\) yields a smaller \(|\mathrm{std}|\); a zero coefficient with positive \(v\) and \(p\) is exactly zero; \(q=0\), \(v=0\), and \(a\ge 0\) fail closed; a non-event clock and an overflowing product fail closed. - Driver et al. (2017, p. 16 `TIPREDEFFECTstd`; §7.2; Eq. 3; footnote 4; JSS PDF re-opened 2026-08-23T16:21Z) recovers a known standardised continuous TI effect \(B\cdot\sqrt{v}/\sqrt{-q/(2a)}\) at machine-scale RMSE, and that RMSE is smaller than treating unstandardised \(B\), asymptotic \((-B/a)\cdot\sqrt{v}/\sqrt{p}\), finite-interval \(A^{-1}[e^{A\Delta t}-I]B\cdot\sqrt{v}/\sqrt{p}\), or \(B\cdot\sqrt{v}/\sqrt{\mathrm{trait}+p+\mathrm{added}}\) as `TIPREDEFFECTstd`; a larger positive \(q\) yields a smaller \(|\mathrm{std}|\); a zero coefficient with positive \(v\) and \(p\) is exactly zero; \(q=0\), \(v=0\), and \(a\ge 0\) fail closed; a non-event clock and an overflowing product fail closed.