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/chooseNumBins.m

View on GitHub
function [NumBins, Info] = chooseNumBins(X, ContinuousMask, NEff, Options)
%chooseNumBins Data-driven bin count for fscchi2 / fsrftest continuous binning.
%
%   [NumBins, Info] = chooseNumBins(X, ContinuousMask, NEff) returns a single
%   positive-integer NumBins to pass to fscchi2 / fsrftest, chosen from the data
%   rather than left at the fixed default of 10. Those filters bin ONLY the
%   continuous predictors before the chi-square / F-test, so the bin count trades
%   resolution (more bins) against contingency-cell reliability (fewer bins);
%   a fixed 10 ignores both the sample size and the spread of the columns.
%
%   Method. For each continuous predictor we compute the Freedman-Diaconis bin
%   count -- range / (2 * IQR * n^(-1/3)) -- which is IQR-based and therefore
%   robust to the skew and outliers a gencfeatures pool tends to produce (unlike
%   Sturges' rule, which assumes near-normal data, or Scott's rule, whose use of
%   the standard deviation is inflated by outliers). We take the MEDIAN of those
%   per-column counts (robust to a few wild columns), then apply two guards:
%     * a validity cap floor(NEff / MinPerCell) so the sparsest cells of the
%       (bin x class) contingency table stay populated enough for the chi-square
%       approximation to hold (Cochran's rule: expected cell count >= 5). The
%       caller passes NEff already reduced to the constraint that binds -- the
%       minority-class count for classification (the fewest-observations row of
%       the table, whose expected cell count under independence is ~NEff/NumBins)
%       and the row count for regression -- so the per-cell cap is exactly
%       floor(NEff / MinPerCell) with MinPerCell = 5, no separate class factor.
%     * a floor of MinBins (default 5) so the test keeps some resolution.
%   Columns with zero IQR (constant / near-constant) contribute no FD estimate.
%
%   Inputs:
%     X              - table or numeric matrix of predictors
%     ContinuousMask - (1,p) logical, true for continuous predictors to bin
%                      (categoricals are excluded; they are not binned)
%     NEff           - (1,1) effective sample size, ALREADY reduced to the binding
%                      constraint: minority-class count for classification, row
%                      count for regression
%     Options.MinPerCell - (1,1) target minimum expected count in the sparsest
%                      contingency cell (default 5, per Cochran's rule)
%     Options.MinBins    - (1,1) lower floor on the returned count (default 5)
%
%   Outputs:
%     NumBins - (1,1) positive integer bin count for fscchi2 / fsrftest
%     Info    - struct with fields:
%                 .FDBins       - (1,:) per-column Freedman-Diaconis counts used
%                 .MedianFDBins - (1,1) median before guards
%                 .ValidityCap  - (1,1) floor(NEff / MinPerCell)
%                 .Reasoning    - (1,1) string explanation
%
%   References:
%     Freedman D, Diaconis P. "On the histogram as a density estimator: L2
%     theory." Zeitschrift fur Wahrscheinlichkeitstheorie und verwandte Gebiete,
%     57(4):453-476, 1981. (bin width = 2 * IQR * n^(-1/3))
%     Cochran WG. "Some methods for strengthening the common chi-square tests."
%     Biometrics 10(4):417-451, 1954. (expected cell count >= 5 for validity)

% Copyright 2026 The MathWorks, Inc.

    arguments
        X
        ContinuousMask (1,:) logical
        NEff (1,1) double {mustBePositive}
        Options.MinPerCell (1,1) double {mustBePositive} = 5
        Options.MinBins (1,1) double {mustBePositive, mustBeInteger} = 5
    end

    if istable(X)
        X = X{:, ContinuousMask};
    else
        X = X(:, ContinuousMask);
    end
    X = double(X);

    ValidityCap = max(Options.MinBins, floor(NEff / Options.MinPerCell));

    % Per-column Freedman-Diaconis bin counts over the continuous predictors.
    n = size(X, 1);
    FDBins = double.empty(1, 0);
    for c = 1:size(X, 2)
        Col = X(~isnan(X(:, c)), c);
        if numel(Col) < 2
            continue;
        end
        Iqr = iqr(Col);
        Rng = max(Col) - min(Col);
        if Iqr <= 0 || Rng <= 0
            continue;   % constant / near-constant column contributes no estimate
        end
        Width = 2 * Iqr * n^(-1/3);
        FDBins(end+1) = max(1, ceil(Rng / Width)); %#ok<AGROW>
    end

    if isempty(FDBins)
        % No informative continuous column (all categorical or constant): the
        % bin count is irrelevant to the ranking, so return the floor.
        MedianFDBins = Options.MinBins;
    else
        MedianFDBins = median(FDBins);
    end

    NumBins = max(Options.MinBins, min(round(MedianFDBins), ValidityCap));

    Info = struct( ...
        FDBins = FDBins, ...
        MedianFDBins = MedianFDBins, ...
        ValidityCap = ValidityCap, ...
        Reasoning = sprintf(...
            "Freedman-Diaconis median=%.0f over %d continuous col(s), " + ...
            "capped to [%d, %d] (NEff=%d, >=%d/cell) -> %d bins", ...
            MedianFDBins, numel(FDBins), Options.MinBins, ValidityCap, ...
            NEff, Options.MinPerCell, NumBins));
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: 468kUpdated Claude CodeCodexUnverifiedInstall
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: 319kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 319kUpdated 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: 319kUpdated Claude CodeCodexVerifiedInstall