Elamur Groupantsmlab-ext-nist-datasets
The groupant beside mlab/tests/ext_nist.rs in the Mova repository: what the code computes, what it never does, how to go back from its outputs, its invariants, cost and checks.
Facts
Implementation material. A Mova code model describes this code at one source revision: partial models, finite checks and their stated scope. It is not a reviewed general exposition, and its labels are shown as written, never raised.
Components
| component | label | where |
|---|---|---|
| approximant | partial | approximant |
| ast | partial | ../syntax.json |
| cost | symbolic source derived | invariant |
| descriptant | present | descriptant |
| invariant | partial | invariant |
| tests | partial bounded parent evidence | tests |
Obernants
The typed relations around the forward model: the behaviour complement (what the code never does), the reverse reconstruction (every input compatible with an output) and specialised views.
| id | type | label | where |
|---|---|---|---|
behavior_complement | behavior_complement | partial source derived model | behavior_complement |
reverse_reconstruction | reverse_reconstruction | partial source derived model | reverse_reconstruction |
design_transport | design_transport | partial source derived model | design_transport |
effect_boundary | effect_boundary | partial source derived model | effect_boundary |
branch_relation | branch_relation | partial source derived model | branch_relation |
resource_cost | resource_cost | partial source derived model | resource_cost |
Model text
descriptant
mlab-ext-nist-datasets
lls_case builds intercept/polynomial/multi-predictor design and checks fitlm plus backslash output length; ANOVA expects9 outputs and exact df; univ computes autocorrelation. suite aggregates threshold failures; detail env affects printing only.
Epoch 2026-10-31: isolated source-bound fragment over bytes [3002,11615); includes no claim beyond this relation. Source hash 82ac1127de001872692aa31cc92c6bf01f367afe18cfbd08941641a96851256c.
approximant
datasets: forward relation
lls_case builds intercept/polynomial/multi-predictor design and checks fitlm plus backslash output length; ANOVA expects9 outputs and exact df; univ computes autocorrelation. suite aggregates threshold failures; detail env affects printing only.
Domain: Well-formed one-based header spans and bounded synthetic data; frozen certified datasets unexecuted.
Executable surrogate gap: no measured cheaper implementation with a useful scoped metric is supplied. Categorical/error/effect relations are exact obligations, not percent losses. Future numeric caps only1/2/5/10%.
invariant
datasets: invariant and cost
One predictor degree depends on count of certified coefficients and B0; multiple predictors use original columns; no-intercept starts power1.
read accesses fixed .dat paths; no selected test calls read, lls_case, anova_case, univ_case or dataset suite.
Scope-specific cost: OLS dataset routes O(n*p²+p³) through stats plus independent solve; ANOVA/univariate O(N), report O(number of compared quantities). These costs are source-derived, not measured.
models
behavior_complement
Synthetic helper success does not certify Norris/Filip/Longley/ANOVA or UNIV thresholds. File-dependent negative controls remain unexecuted.
reverse_reconstruction
Reported LRE rows admit many generated data sets and numerical errors; thresholds determine accepted region not unique reference values.
design_transport
One predictor degree depends on count of certified coefficients and B0; multiple predictors use original columns; no-intercept starts power1.
effect_boundary
read accesses fixed .dat paths; no selected test calls read, lls_case, anova_case, univ_case or dataset suite.
branch_relation
lls_case builds intercept/polynomial/multi-predictor design and checks fitlm plus backslash output length; ANOVA expects9 outputs and exact df; univ computes autocorrelation. suite aggregates threshold failures; detail env affects printing only.
resource_cost
OLS dataset routes O(n*p²+p³) through stats plus independent solve; ANOVA/univariate O(N), report O(number of compared quantities). These costs are source-derived, not measured.
tests
Evidence boundary
See ../tests.md for the actual selected source test and mutant. These tests sample the file, not every fragment branch. Synthetic helper success does not certify Norris/Filip/Longley/ANOVA or UNIV thresholds. File-dependent negative controls remain unexecuted. Other branches here are source-derived and unexecuted unless the parent test explicitly calls them.
This page in Markdown → https://md.groupants.elamur.ai/free/en/mova/mlab/tests/ext_nist.rs.groupant/datasets.groupant/