wvc_select — WVC — Weighted Variable Contribution

Group: Variable selector · Registry tolerance: 0.7

Description

WVC weighted-variable-component top-k

From the pls4all.sklearn.WVCSelector docstring:

WVC-PLS — weighted variable contribution top-k selector.

Registry note — R plsVarSel::WVC_pls top-k weighted-variable scoring. Mask RMSE-rel ~0=perfect, ~1=half disagree, ~1.41=disjoint; tolerance 0.7 enforces ~50% overlap.

Parameters

Name

Type

Default

Notes

top_k

int

None

Number of features to retain.

n_components

int

2

Number of latent components extracted (k).

normalize

bool

True

Normalize per-variable scores to sum to one before ranking.

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.

Mathematical principle

WVC builds a normalised weighted-variable-contribution score from the SVD of the PLS components. Each feature’s contribution to each component is weighted by the component’s singular value (importance), then summed and normalised. Sort by WVC, take the top-\(k\).

Compared to VIP, WVC uses SVD-based weighting which downweights components dominated by noise; this gives more stable rankings when \(k\) is over-specified.

Implementation

n4m_wvc_select.

MATLAB header (bindings/matlab/+pls4all/wvc_select.m):

pls4all.wvc_select  Weighted-vector correlation top-k selector.

Usage

Every pls4all binding tab dispatches into the same C kernel; the external libraries listed at the bottom of the page are the parity references registered in benchmarks.parity_timing.registry. Switch tabs to read the same fit in your language. The R package now ships drop-in-compatible facades for the CRAN pls package (plsr, pcr, mvr) and for the mdatools::pls(x, y, ...) matrix idiom — those tabs appear only on the methods that have a meaningful equivalence.

pls4all bindings

/* C ABI — libn4m */
n4m_context_t* ctx = n4m_context_create();
n4m_config_t*  cfg = n4m_config_create();
n4m_method_result_t* res = NULL;
n4m_wvc_select_fit(ctx, cfg, &x_view, &y_view, /* hyperparams */, &res);
/* … read coefficients / mask / scores via */
/* n4m_method_result_get_double_matrix / vector / scalar … */
n4m_method_result_destroy(res);
n4m_config_destroy(cfg);
n4m_context_destroy(ctx);

import pls4all
from pls4all._methods import wvc_select_fit
with pls4all.Context() as ctx, pls4all.Config() as cfg:
    res = wvc_select_fit(ctx, cfg, X, y, n_components=4)
# then: res.matrix("predictions"), res.matrix("coefficients"),
# res.vector("mask"), res.scalar("intercept"), …

from pls4all.sklearn import WVCSelector
mdl = WVCSelector(top_k, n_components=2, normalize=True)
mdl.fit(X, y)
y_hat = mdl.predict(X_test)

library(pls4all)
# Unified low-level dispatcher (May 2026 R cleanup):
res <- pls4all_method("wvc_select", X, y,
                      n_components = 4L, params = list(top_k = 10L))
# res is a named list with MethodResult arrays/scalars.
# selected_indices / top_k_intervals are 1-based.

res = pls4all.wvc_select(X, y, 4);
% see header of bindings/matlab/+pls4all/wvc_select.m for full
% parameter surface:
%   res = wvc_select(X, Y, n_components, top_k, normalize)
yhat = predict(res, Xtest);

No idiomatic classdef wrapper — invoke pls4all.fit("wvc_select", X, y, …) directly from the unified MEX factory.

Registry parity references 📐

  • 📐 ref.r_plsvarsel (R · r) — plsVarSel 0.10.0 · qualitative (rmse_rel ≤ 7e-01) — R plsVarSel::WVC_pls — weighted-variable-component scoring.

Benchmarks

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 = pls4all binding 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: qualitative — shape/smoke comparison only. The external library and pls4all do not produce numerically equivalent output for this method (see the MethodSpec notes); the rmse_rel_tol ≤ 7e-01 budget is set wide on purpose. Treat ~ shape as “we ran both, both finished”, not as numerical agreement.

Rows tagged with 📐 are the canonical parity references for this method (declared in parity_timing.registry). C++ and external rows show reference parity; pls4all language bindings show binding parity against the C++ backend. Hover the icon for role and tolerance band.

BackendParity200×40 (ms)
C++ native · libn4m
pls4all.cpp.blas+omp✓ J 1.001.49 ms
Python · pls4all
pls4all.python✓ J 1.001.48 ms🏆
pls4all.sklearn✓ J 1.001.63 ms
R · pls4all
pls4all.R✓ J 1.004.23 ms
pls4all.R.formula✓ J 1.004.97 ms
pls4all.R.mdatools✓ J 1.005.30 ms
pls4all.R.pls✓ J 1.005.72 ms
R · external
📐ref.r_plsvarselsource14.4 ms
BackendParity200×40 (ms)
C++ native · libn4m
pls4all.cpp.blas+omp✓ J 1.001.55 ms
Python · pls4all
pls4all.python✓ J 1.001.53 ms🏆
pls4all.sklearn✓ J 1.001.65 ms
R · pls4all
pls4all.R✓ J 1.004.24 ms
pls4all.R.formula✓ J 1.004.92 ms
pls4all.R.mdatools✓ J 1.005.18 ms
pls4all.R.pls✓ J 1.005.28 ms
R · external
📐ref.r_plsvarselsource13.0 ms
BackendParity200×40 (ms)
C++ native · libn4m
pls4all.cpp.blas+omp✓ J 1.001.50 ms🏆
Python · pls4all
pls4all.python✓ J 1.002.65 ms
pls4all.sklearn✓ J 1.001.68 ms
R · pls4all
pls4all.R✓ J 1.004.88 ms
pls4all.R.formula✓ J 1.006.13 ms
pls4all.R.mdatools✓ J 1.005.79 ms
pls4all.R.pls✓ J 1.005.46 ms
R · external
📐ref.r_plsvarselsource13.8 ms

See also: benchmark overview · methods index · interactive dashboard