Release v9.1.0 — extended analysis, visualization, and validation

New modules: VolcanoPlot, ANOVAStats, KruskalStats, lda, plsr, ica, hca,
MLClassifier, SpectralAutoencoder, SpectralMLP, SpectralCNN, ModelValidator,
TrainTestSplit, CrossValidator, SpectraPlot, FeatureHeatmap, CorrelationMatrix,
PValueTable, Biplot, CoefficientPlot, FeatureImportancePlot, FoldChangePlot,
GroupComparison, MetaboliteDashboard.

C backend: nipals_full, scale_transform, xcorr_max_shift, pqn_median_quotient
kernels added to _native_backend.c. Fixed reference leak (O→N in Py_BuildValue).

Docs: feedback/issue reporter page, skeleton loader, download stats strip,
topbar light-theme fix, GitHub Actions path filter updated.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Author: aeiwz

.github/workflows/docs.yml

@@ -4,13 +4,11 @@ on:
   pull_request:
     paths:
       - "docs/**"
-      - "metbit/**"
       - ".github/workflows/docs.yml"
   push:
     branches: [main]
     paths:
       - "docs/**"
-      - "metbit/**"
       - ".github/workflows/docs.yml"
 
 permissions:

docs/app/docs/api/lazy_opls_da/page.mdx

@@ -27,9 +27,19 @@ Parameters:
 <details>
 <summary><code>__init__(data, groups, working_dir, feature_, n_components, scaling, estimator, kfold, random_state, auto_ncomp, permutation, n_permutation, n_jobs, VIP, VIP_threshold, linear_regression, FC_threshold, p_val_threshold)</code></summary>
 
+This function takes in a dataframe and a list of y values and returns the project_name model.
+
+##### Parameters
+
+- `data`: pandas dataframe The dataframe to be used.
+- `y`: list The list of y values.
+- `n_components`: int The number of components to use. lazy_opls_da(data, y, n_components).fit()
+
 </details>
 
 <details>
 <summary><code>fit(marker_color, custom_color, custom_shape, symbol_dict, custom_legend_name, marker_label, marker_size, marker_opacity, individual_ellipse)</code></summary>
 
+Fit the OPLS-DA model to all pairwise group comparisons and save plots and data.
+
 </details>

docs/app/docs/api/page.mdx

@@ -1,47 +1,58 @@
 # API Reference
 
-Browse the main public APIs exposed by `metbit`. The preferred import style is:
+Browse the public API exposed by `metbit` (sorted by category).
 
-```python
-from metbit import pca, opls_da, nmr_preprocessing, Normalization
-```
+## Data Processing
 
-Advanced users can also import from subpackages such as `metbit.analysis`, `metbit.nmr`, `metbit.preprocessing`, `metbit.stats`, and `metbit.apps`.
-
-## NMR Data Processing
-
-- [nmr_preprocessing](/docs/api/nmr_preprocess) - read Bruker FID folders and generate processed spectra.
-- [calibrate](/docs/api/calibrate) - calibrate chemical-shift axes.
-- [baseline_correct and bline](/docs/api/baseline) - baseline correction helpers.
-- [detect_multiplets, icoshift_align, and PeakAligner](/docs/api/peak_processe) - peak detection and interval alignment utilities.
-- [peak_chops](/docs/api/peak_processe) - peak-region extraction.
-- [denoise_spec](/docs/api/denoise_spec) - spectral denoising helpers.
-
-## Normalization and Pretreatment
-
-- [Normalization](/docs/api/spec_norm) - static PQN, SNV, MSC, and combined normalization methods.
-- [Normalise](/docs/api/utility) - class-based normalization utilities.
-- [scaler](/docs/api/scaler) - scaling utilities used by statistical models.
-- [pretreatment](/docs/api/pretreatment) - preprocessing and transformation helpers.
+- [calibrate](/docs/api/metbit.nmr.calibrate)
+- [nmr_preprocessing](/docs/api/metbit.nmr.preprocess)
+- [Normalise](/docs/api/metbit.stats.normalise)
+- [Normalization](/docs/api/metbit.preprocessing.normalize)
+- [peak_chops](/docs/api/metbit.nmr.peaks)
 
 ## Statistical Models
 
-- [pca](/docs/api/metbit) - PCA model with interactive scores, loading, variance, and trajectory plots.
-- [opls_da](/docs/api/metbit) - OPLS-DA model with cross-validation, permutation testing, VIP, loading, S-plot, and scores plotting.
-- [lazy_opls_da](/docs/api/lazy_opls_da) - convenience workflow for OPLS-DA analysis.
-- [OPLS](/docs/api/opls) - lower-level OPLS model implementation.
-- [PLS](/docs/api/pls) - lower-level PLS model implementation.
-- [CrossValidation](/docs/api/cross_validation) - cross-validation helpers.
-- [VIP helpers](/docs/api/vip) - Variable Importance in Projection utilities.
-- [UnivarStats](/docs/api/utility) - univariate statistical analysis helpers.
-
-## Visualization and Interactive Apps
-
-- [STOCSY](/docs/api/STOCSY) - Statistical Total Correlation Spectroscopy.
-- [STOCSY_app](/docs/api/ui_stocsy) - interactive Dash STOCSY app.
-- [pickie_peak](/docs/api/ui_picky_peak) - interactive peak-picking app.
-- [annotate_peak](/docs/api/annotate_peak) - peak annotation helper.
-- [get_intensity](/docs/api/take_intensity) - peak intensity extraction.
-- [plotting](/docs/api/plotting) - reusable plotting helpers.
-- [boxplot](/docs/api/boxplot) - boxplot visualization utilities.
-- [pca_ellipse](/docs/api/pca_ellipse) - confidence ellipse plotting.
+- [opls_da](/docs/api/metbit.analysis.opls_da)
+- [pca](/docs/api/metbit.analysis.pca)
+- [UnivarStats](/docs/api/metbit.stats.univariate)
+
+## Data Visualization
+
+- [STOCSY](/docs/api/metbit.analysis.stocsy)
+
+## Other
+
+- [ANOVAStats](/docs/api/metbit.stats.multitest)
+- [available_cv_strategies](/docs/api/metbit.validation.splitter)
+- [backend_info](/docs/api/metbit._native)
+- [baseline_correct](/docs/api/metbit.preprocessing.baseline)
+- [Biplot](/docs/api/metbit.viz.interpretation)
+- [bline](/docs/api/metbit.preprocessing.baseline)
+- [ChunkedSTOCSY](/docs/api/metbit.analysis.large_scale)
+- [CoefficientPlot](/docs/api/metbit.viz.interpretation)
+- [CorrelationMatrix](/docs/api/metbit.viz.summary)
+- [CrossValidator](/docs/api/metbit.validation.splitter)
+- [detect_multiplets](/docs/api/metbit.nmr.alignment)
+- [feature_preselection](/docs/api/metbit.analysis.large_scale)
+- [FeatureHeatmap](/docs/api/metbit.viz.summary)
+- [FeatureImportancePlot](/docs/api/metbit.viz.interpretation)
+- [FoldChangePlot](/docs/api/metbit.viz.profiling)
+- [gpu_available](/docs/api/metbit._native)
+- [GroupComparison](/docs/api/metbit.viz.profiling)
+- [hca](/docs/api/metbit.analysis.multivariate)
+- [ica](/docs/api/metbit.analysis.multivariate)
+- [icoshift_align](/docs/api/metbit.nmr.alignment)
+- [KruskalStats](/docs/api/metbit.stats.multitest)
+- [LargeScaleAlignment](/docs/api/metbit.analysis.large_scale)
+- [lda](/docs/api/metbit.analysis.multivariate)
+- [memory_report](/docs/api/metbit.analysis.large_scale)
+- [MemoryEstimator](/docs/api/metbit.analysis.large_scale)
+- [ModelValidator](/docs/api/metbit.validation.metrics)
+- [native_available](/docs/api/metbit._native)
+- [PeakAligner](/docs/api/metbit.nmr.alignment)
+- [plsr](/docs/api/metbit.analysis.multivariate)
+- [PValueTable](/docs/api/metbit.viz.summary)
+- [SpectraPlot](/docs/api/metbit.viz.spectra)
+- [TrainTestSplit](/docs/api/metbit.validation.splitter)
+- [VolcanoPlot](/docs/api/metbit.stats.multitest)
+

docs/app/docs/api/pretreatment/page.mdx

@@ -6,6 +6,8 @@
 
 ### Scaler 
 
+A scaler to scale data matrix
+
 #### Methods
 
 <details>

metbit/__init__.py

@@ -6,7 +6,7 @@
 __maintainer__ = "aeiwz"
 __status__ = "Development"
 __copyright__ = "Copyright 2024, Theerayut"
-__version__ = "9.0.0"
+__version__ = "9.1.0"
 
 # Core analysis
 from metbit.analysis.opls_da import opls_da
@@ -35,6 +35,28 @@
 from metbit.preprocessing.baseline import baseline_correct, bline
 from metbit.preprocessing.normalize import Normalization
 
+# Extended statistics
+from metbit.stats.multitest import VolcanoPlot, ANOVAStats, KruskalStats
+
+# Extended multivariate analysis
+from metbit.analysis.multivariate import lda, plsr, ica, hca
+
+# Validation
+from metbit.validation.metrics import ModelValidator
+from metbit.validation.splitter import TrainTestSplit, CrossValidator, available_cv_strategies
+
+# Machine learning classifiers (optional: requires xgboost)
+try:
+    from metbit.ml.classifiers import MLClassifier
+except Exception:  # pragma: no cover
+    pass
+
+# Deep learning models (optional: requires torch)
+try:
+    from metbit.dl.models import SpectralAutoencoder, SpectralMLP, SpectralCNN
+except Exception:  # pragma: no cover
+    pass
+
 # Dash apps (optional heavy deps)
 try:
     from metbit.apps.annotate import annotate_peak
@@ -51,6 +73,24 @@
 except Exception:  # pragma: no cover
     pass
 
+# Visualization - spectra
+from metbit.viz.spectra import SpectraPlot
+
+# Visualization - statistical summaries
+from metbit.viz.summary import FeatureHeatmap, CorrelationMatrix, PValueTable
+
+# Visualization - model interpretation
+from metbit.viz.interpretation import Biplot, CoefficientPlot, FeatureImportancePlot
+
+# Visualization - metabolite profiling
+from metbit.viz.profiling import FoldChangePlot, GroupComparison
+
+# Dash app: metabolite dashboard (optional heavy deps)
+try:
+    from metbit.viz.profiling import MetaboliteDashboard
+except Exception:  # pragma: no cover
+    pass
+
 # Legacy flat-layout class not yet moved to a sub-package
 try:
     from metbit.lazy_opls_da import lazy_opls_da
@@ -80,6 +120,38 @@
     "baseline_correct",
     "bline",
     "Normalization",
+    # Extended stats
+    "VolcanoPlot",
+    "ANOVAStats",
+    "KruskalStats",
+    # Extended multivariate
+    "lda",
+    "plsr",
+    "ica",
+    "hca",
+    # Validation
+    "ModelValidator",
+    "TrainTestSplit",
+    "CrossValidator",
+    "available_cv_strategies",
+    # Visualization - spectra
+    "SpectraPlot",
+    # Visualization - statistical summaries
+    "FeatureHeatmap",
+    "CorrelationMatrix",
+    "PValueTable",
+    # Visualization - model interpretation
+    "Biplot",
+    "CoefficientPlot",
+    "FeatureImportancePlot",
+    # Visualization - metabolite profiling
+    "FoldChangePlot",
+    "GroupComparison",
+]
+_optional = [
+    "annotate_peak", "STOCSY_app", "pickie_peak", "lazy_opls_da",
+    "MLClassifier",
+    "SpectralAutoencoder", "SpectralMLP", "SpectralCNN",
+    "MetaboliteDashboard",
 ]
-_optional = ["annotate_peak", "STOCSY_app", "pickie_peak", "lazy_opls_da"]
 __all__ = _always_exported + [n for n in _optional if n in globals()]