sparse_pls_da — Sparse PLS-DA (Lê Cao 2008)

Group: Sparse · Registry tolerance: 1e-08

Description

Sparse PLS-DA (§7)

Registry note — R spls::splsda uses an LDA classifier on PLS scores; pls4all and SparsePlsDaPythonReference use argmax of the regression decision scores. Both emit one-hot predictions; differences appear only at the decision boundary.

Parameters

Name	Type	Default	Notes
n_components	int	4	registry benchmark cell value
sparsity_lambda	float	0.05	registry benchmark cell value
n_classes	int	3	registry benchmark cell value

Explanations

Bibliographic source

Lê Cao, K.-A., Rossouw, D., Robert-Granié, C. & Besse, P. (2008). A sparse PLS for variable selection when integrating omics data. Statistical Applications in Genetics and Molecular Biology 7(1). Verified primary link: https://doi.org/10.2202/1544-6115.1390.

Mathematical principle

Class labels are encoded as one-hot response columns and a sparse PLS regression model produces one score per class. Prediction selects the largest score and writes a one-hot class vector. This supplies a discriminant rule but does not normalize the scores into probabilities.

Appropriate uses

Sparse multiclass discrimination with one-hot responses and an interpretable wavelength subset.

Limits and validation

The prediction surface emits an argmax class and one-hot output, not calibrated softmax probabilities.

Implementation

The C result wrapper performs argmax over class scores and emits zero/one entries. Probability calibration, softmax conversion, and rejection thresholds are not part of this method.

Sources and provenance

Current implementation: cpp/src/c_api/c_api_method_result.cpp.

API and bindings

C ABI (ABI 2): n4m_estimators_sparse_pls_da_fit. Use the linked public header for the exact signature, configuration, and result handles.

Python: no current AST-verified public n4m re-export was found for this method. The linked C ABI above is the documented surface in this checkout.

R (source-verified): sparse_pls_da_fit(X, y_labels, n_components, sparsity_lambda = 0.05).

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

MATLAB / Octave (source-verified): sparse_pls_da(X, y_labels, n_components, sparsity_lambda).

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

Registry parity references 📐

• 📐 ref.python_chun_keles_splsda (python · python) — chun_keles_splsda 1.0 · strict (rmse_rel ≤ 1e-08) — Sparse SIMPLS (Chun & Keles 2010) on dummy-coded class labels, followed by argmax over decision scores. Mirrors pls4all’s n4m_estimators_sparse_pls_da_fit (default, cfg.sparse_simpls_legacy = 0) bit-for-bit.

• 📐 ref.r_spls (R · r) — spls 2.3.2 · strict (rmse_rel ≤ 1e-08) — R spls::splsda (Chun & Keles). Predictions returned as hard class labels by the package; we one-hot encode them to match pls4all’s soft-assignment prediction shape, so the parity check is on the classification boundary rather than continuous score values.

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-08).

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	200×50 (ms)
C++ native · libn4m
pls4all.cpp.blas+omp	✓ ref	7.43 ms
Python · archived pls4all benchmark
pls4all.python	✓ bind	4.50 ms
pls4all.sklearn	⇄ +7e-01	2.45 ms🏆
R · archived pls4all benchmark
pls4all.R	✓ bind	7.20 ms
pls4all.R.formula	✓ bind	8.68 ms
pls4all.R.mdatools	✓ bind	8.43 ms
pls4all.R.pls	✓ bind	8.53 ms
Python · external
📐ref.python_chun_keles_splsda	source	3.58 ms
R · external
📐ref.r_spls	⇄ +1e+00	82.9 ms

3 threads

Backend	Parity	200×50 (ms)
C++ native · libn4m
pls4all.cpp.blas+omp	✓ ref	12.7 ms
Python · archived pls4all benchmark
pls4all.python	✓ bind	8.19 ms
pls4all.sklearn	⇄ +7e-01	5.55 ms🏆
R · archived pls4all benchmark
pls4all.R	✓ bind	23.8 ms
pls4all.R.formula	✓ bind	31.7 ms
pls4all.R.mdatools	✓ bind	37.4 ms
pls4all.R.pls	✓ bind	15.7 ms
Python · external
📐ref.python_chun_keles_splsda	source	14.4 ms
R · external
📐ref.r_spls	⇄ +1e+00	117.9 ms

10 threads

Backend	Parity	200×50 (ms)
C++ native · libn4m
pls4all.cpp.blas+omp	✓ ref	7.83 ms
Python · archived pls4all benchmark
pls4all.python	✓ bind	7.59 ms
pls4all.sklearn	⇄ +7e-01	3.94 ms🏆
R · archived pls4all benchmark
pls4all.R	✓ bind	12.2 ms
pls4all.R.formula	✓ bind	47.2 ms
pls4all.R.mdatools	✓ bind	26.4 ms
pls4all.R.pls	✓ bind	32.6 ms
Python · external
📐ref.python_chun_keles_splsda	source	6.94 ms
R · external
📐ref.r_spls	⇄ +1e+00	91.2 ms

---

See also: benchmark overview · methods index · interactive dashboard