wvc_threshold_select — WVC-threshold selection

Group: Variable selector · Registry tolerance: 1e-06

Description

WVC threshold-based selection (§18 Phase 5r)

Registry note — R plsVarSel::WVC_pls with explicit median-scaled threshold. Default _wvc_threshold_select_pls4all path mirrors the same R call, giving bit-exact mask parity. The C++ min-selected kernel is opt-in via legacy=True.

Parameters

Name	Type	Default	Notes
n_components	int	2	current public binding signature
normalize	bool	True	current public binding signature
score_threshold	float	0.0	current public binding signature
threshold_factor	float	1.0	current public binding signature
min_selected	int | None	None	current public binding signature

Explanations

Bibliographic source

Andries, J. P. M. & Vander Heyden, Y. (2011). Improved variable reduction in partial least squares modelling based on predictive-property-ranked variables and adaptation of partial least squares complexity. Analytica Chimica Acta 705(1–2), 292–305. https://doi.org/10.1016/j.aca.2011.06.037 — same paper as wvc_select; introduces both the top-\(k\) ranking and the threshold / factor-of-mean rules used here.

Mathematical principle

After WVC scores are computed, the implementation uses one cutoff: the larger of score_threshold and threshold_factor times the mean score. Variables meeting it are retained, with the largest scores added if fewer than min_selected survive.

Appropriate uses

Applying an explicit cutoff to WVC scores for a compact selected set.

Limits and validation

One cutoff max(score_threshold, threshold_factor*mean_score) is used, with a min_selected fallback; no competing-rule CV selection is performed.

Implementation

The routine does not compare two thresholding rules by cross-validation. Any predictive validation of the chosen cutoff must be performed by the caller.

Sources and provenance

Current implementation: cpp/src/core/wvc_selection.cpp.

API and bindings

C ABI (ABI 2): n4m_feature_selection_wvc_threshold_select. Use the linked public header for the exact signature, configuration, and result handles.

Python (verified public re-export):

from n4m.feature_selection.wrapper import WVCThreshold

Source signature: WVCThreshold(*, n_components: int = 2, normalize: bool = True, score_threshold: float = 0.0, threshold_factor: float = 1.0, min_selected: int | None = None) (n4m/_impl/selection.py).

R (source-verified): wvc_threshold_select(X, Y, n_components, normalize = TRUE, threshold = 0.0, threshold_factor = 1.0, min_selected = 1L).

The source signature has additional required inputs, so no example call is fabricated.

MATLAB / Octave (source-verified): wvc_threshold_select(X, Y, n_components, normalize, threshold, ... threshold_factor, min_selected).

The source signature has additional required inputs, so no example call is fabricated.

Registry parity references 📐

• 📐 ref.r_plsvarsel (R · r) — plsVarSel 0.10.0 · strict (rmse_rel ≤ 1e-06) — R plsVarSel::WVC_pls with explicit threshold — picks features whose weighted-variable scores exceed the median × threshold-factor.

Benchmarks

Archived measurement identity. Backend labels in this table are the raw IDs recorded when the benchmark ran (including historical pls4all.* IDs). They preserve measurement provenance and do not describe a current public Python, R, or MATLAB binding; use the source-verified API and bindings section above for current entry points.

Adaptive wall-clock per cell measured against full_matrix.csv. Only backends that implement this method are listed; libraries without the method are omitted.

Verdict  ·  ✓ ref / ≈ ref / ~ shape mark a reference-gate pass at strict / relaxed / qualitative tolerance  ·  ✓ bind = archived binding-harness result agrees with the C++ baseline  ·  ⇄ cross-check = documented by-design selector/RNG/model, noncanonical API/facade convention, or secondary oracle  ·  ✗ divergent  ·  ⚠ error  ·  — not run. The fastest backend per column is marked 🏆.

Reference gate: strict — numeric equivalence (rmse_rel_tol ≤ 1e-06).

Rows tagged with 📐 are the canonical parity references for this method (declared in parity_timing.registry). C++ and external rows show reference parity; archived language-harness rows show binding parity against the C++ backend. Hover the icon for role and tolerance band.

1 thread

Backend	Parity	200×40 (ms)
C++ native · libn4m
pls4all.cpp.blas+omp	✓ J 1.00	336.1 ms
Python · archived pls4all benchmark
pls4all.python	✓ J 1.00	344.7 ms
pls4all.sklearn	⇄ J 0.69	1.67 ms🏆
R · archived pls4all benchmark
pls4all.R	⇄ J 0.69	4.07 ms
pls4all.R.formula	⇄ J 0.69	4.66 ms
pls4all.R.mdatools	⇄ J 0.69	4.59 ms
pls4all.R.pls	⇄ J 0.69	5.51 ms
R · external
📐ref.r_plsvarsel	source	13.0 ms

3 threads

Backend	Parity	200×40 (ms)
C++ native · libn4m
pls4all.cpp.blas+omp	✓ J 1.00	304.2 ms
Python · archived pls4all benchmark
pls4all.python	✓ J 1.00	301.0 ms
pls4all.sklearn	⇄ J 0.69	1.57 ms🏆
R · archived pls4all benchmark
pls4all.R	⇄ J 0.69	4.31 ms
pls4all.R.formula	⇄ J 0.69	4.42 ms
pls4all.R.mdatools	⇄ J 0.69	4.35 ms
pls4all.R.pls	⇄ J 0.69	4.40 ms
R · external
📐ref.r_plsvarsel	source	12.7 ms

10 threads

Backend	Parity	200×40 (ms)
C++ native · libn4m
pls4all.cpp.blas+omp	✓ J 1.00	340.2 ms
Python · archived pls4all benchmark
pls4all.python	✓ J 1.00	329.0 ms
pls4all.sklearn	⇄ J 0.69	1.61 ms🏆
R · archived pls4all benchmark
pls4all.R	⇄ J 0.69	4.06 ms
pls4all.R.formula	⇄ J 0.69	4.70 ms
pls4all.R.mdatools	⇄ J 0.69	5.14 ms
pls4all.R.pls	⇄ J 0.69	5.45 ms
R · external
📐ref.r_plsvarsel	source	13.5 ms

---

See also: benchmark overview · methods index · interactive dashboard