Skip to content

matlab/matlab-agentic-toolkit / skills-catalog/ai-and-statistics/matlab-engineer-tabular-features

matlab-engineer-tabular-features

Unverified

Use when engineering or selecting the best features for single-response classification or regression in MATLAB, whatever the data's modality — for non-tabular data it routes extraction to a domain skill, then selects, assesses, and delivers on the resulting table. Not for multi-response problems, model training, or raw data acquisition.

FileKindSize
SKILL.mdskill md18.9 KB
manifest.yamlconfig570 B
references/assess.mdmarkdown16.7 KB
references/deliver.mdmarkdown13.0 KB
references/domain-routing.mdmarkdown18.1 KB
references/feature-pool.mdmarkdown10.9 KB
references/intake.mdmarkdown16.9 KB
references/internal-helpers.mdmarkdown12.2 KB
references/provider-protocol.mdmarkdown12.3 KB
references/report.mdmarkdown9.7 KB
references/reproduce-pool.mdmarkdown7.1 KB
references/select.mdmarkdown7.1 KB
scripts/BordaVoter.mother2.3 KB
scripts/ConsensusVoter.mother2.5 KB
scripts/MajorityVoter.mother2.0 KB
scripts/RRFVoter.mother1.9 KB
scripts/assessGenerationStability.mother11.9 KB
scripts/assessKFold.mother20.6 KB
scripts/assessSelectionStability.mother9.8 KB
scripts/baselineComparison.mother18.9 KB
scripts/buildRecipe.mother8.2 KB
scripts/chooseNumBins.mother5.0 KB
scripts/chooseStandardization.mother4.5 KB
scripts/computeNSelect.mother2.9 KB
scripts/createConsensusVoter.mother2.2 KB
scripts/describeFeatures.mother5.1 KB
scripts/expandBlockMap.pbinary440 B
scripts/featureSetQuality.mother12.3 KB
scripts/generateFeatureReport.mother67.2 KB
scripts/generateFeatureReportPdf.mother64.9 KB
scripts/generateFeatures.mother14.6 KB
scripts/hasReportGenerator.mother630 B
scripts/isDomainRecipe.mother1.4 KB
scripts/isStringEmpty.mother577 B
scripts/lambdaSequence.pbinary637 B
scripts/lassoFeatureRanking.pbinary971 B
scripts/macroF1Score.mother931 B
scripts/mrmrFeatureRanking.pbinary239 B
scripts/mutualInformationPanel.pbinary1.1 KB
scripts/ncaFeatureRanking.pbinary464 B
scripts/nogueiraStability.mother3.8 KB
scripts/oobFeatureRanking.pbinary414 B
scripts/plotSelectionConsensus.mother2.7 KB
scripts/plotSelectionDecision.mother6.3 KB
scripts/plotSelectionStability.mother10.2 KB
scripts/plotValidationBaseline.mother6.1 KB
scripts/printSummary.mother18.3 KB
scripts/profileForSplit.mother6.2 KB
scripts/rankAgreement.mother1.3 KB
scripts/rankFeatures.mother6.2 KB
scripts/reserveHoldoutForUser.mother5.5 KB
scripts/runConsensusSelection.mother9.4 KB
scripts/sampleSkewness.mother600 B
scripts/saveResults.mother2.1 KB
scripts/scoreFeaturePanel.pbinary1.6 KB
scripts/screenPredictors.mother13.3 KB
scripts/selectFeatures.mother2.3 KB
scripts/splitStrategy.mother5.4 KB
scripts/stratifiedDownsampleRows.mother2.2 KB
scripts/stratifiedSubsampleIndices.mother3.4 KB
scripts/transformFeatures.mother6.1 KB
scripts/validateRecipe.mother12.3 KB
scripts/writeInferenceScript.mother10.5 KB

scripts/nogueiraStability.m

View on GitHub
function [Stability, Info] = nogueiraStability(Z)
%nogueiraStability Chance-corrected selection-stability index (Nogueira 2018).
%
%   [Stability, Info] = nogueiraStability(Z) measures how consistently a feature
%   selector picks the same features across repeated runs (e.g. k folds). Z is a
%   K-by-p binary matrix over a FIXED shared feature universe: Z(i,j) = 1 if
%   feature j was selected in run i, else 0. K is the number of runs (folds),
%   p the number of candidate features.
%
%   Unlike Kuncheva's index or a plain Jaccard, this estimator does NOT require
%   an equal number of features selected per run, so it handles the variable
%   per-fold subset size that elbow-cut selection (computeNSelect) produces. It
%   scores selection DECISIONS, not feature VALUES, so it is feature-type
%   agnostic (numeric/categorical/continuous all work).
%
%   Definition (see Reference below, eq. 2):
%     Let p_f = mean(Z(:,f)) be feature f's selection frequency, and
%     s_f^2 = (K/(K-1)) p_f (1 - p_f) its unbiased Bernoulli sample variance.
%     Let kbar = mean row-sum (average subset size). Then
%       Stability = 1 - ( (1/p) sum_f s_f^2 ) / ( (kbar/p)(1 - kbar/p) ).
%   The denominator is the variance expected if the same number of features were
%   chosen uniformly at random, so the index is chance-corrected: ~0 means
%   selections are no more consistent than random, 1 means identical every run.
%   It can be slightly negative (below-chance agreement); that is reported
%   honestly rather than clamped.
%
%   Inputs:
%     Z - (K,p) binary (0/1) selection matrix; K >= 2 runs, p >= 1 features
%
%   Outputs:
%     Stability - (1,1) double, the stability index (typically in [0,1], may be
%                 slightly negative; 1 = perfectly stable)
%     Info      - struct with fields:
%                   .SelectionFrequency - (1,p) per-feature selection frequency
%                                         (the "consensus core" companion read)
%                   .MeanSubsetSize     - (1,1) average number selected per run
%                   .NumRuns            - (1,1) K
%                   .NumFeatures        - (1,1) p
%
%   Reference:
%     Nogueira S, Sechidis K, Brown G. "On the Stability of Feature Selection
%     Algorithms." Journal of Machine Learning Research, 18(174):1-54, 2018.
%     https://jmlr.org/papers/v18/17-514.html  (estimator: eq. 2, Sec. 2)

% Copyright 2026 The MathWorks, Inc.

    arguments
        Z {mustBeNumericOrLogical, mustBeReal}
    end

    Z = double(Z);
    if ~ismatrix(Z)
        error('nogueiraStability:notMatrix', 'Z must be a K-by-p matrix.');
    end
    if ~all(ismember(Z(:), [0 1]))
        error('nogueiraStability:notBinary', 'Z must contain only 0 and 1.');
    end

    [K, p] = size(Z);
    if K < 2
        error('nogueiraStability:tooFewRuns', ...
            'Stability needs at least 2 runs (rows); got %d.', K);
    end
    if p < 1
        error('nogueiraStability:noFeatures', 'Z must have at least one column.');
    end

    SelectionFrequency = mean(Z, 1);                 % p_f, one per feature (1,p)
    MeanSubsetSize = mean(sum(Z, 2));                % kbar, average subset size

    % Unbiased per-feature Bernoulli variance, averaged over features.
    UnbiasedVar = (K / (K - 1)) .* SelectionFrequency .* (1 - SelectionFrequency);
    MeanVar = mean(UnbiasedVar);

    % Chance-level variance from the average subset size.
    ExpectedFrac = MeanSubsetSize / p;
    NullVar = ExpectedFrac * (1 - ExpectedFrac);

    if NullVar == 0
        % Every run selected all features (or none): selections are trivially
        % identical, so stability is perfect by convention.
        Stability = 1;
    else
        Stability = 1 - MeanVar / NullVar;
    end

    Info = struct( ...
        SelectionFrequency = SelectionFrequency, ...
        MeanSubsetSize = MeanSubsetSize, ...
        NumRuns = K, ...
        NumFeatures = p);
end
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 469kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 321kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 321kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 321kUpdated Claude CodeCodexVerifiedInstall