columna API reference - v0.3.0
    Preparing search index...

    Interface OlsResult

    interface OlsResult {
        aic: number;
        aicc: number;
        anova: { error: AnovaRow; regression: AnovaRow; total: AnovaRow };
        bic: number;
        coefficients: Coefficient[];
        confidence: number;
        cooksD: Float64Array;
        covariance: Matrix;
        design: Matrix;
        dffits: Float64Array;
        durbinWatson: number;
        fitted: Float64Array;
        intercept: boolean;
        leverage: Float64Array;
        logLik: number;
        n: number;
        names: string[];
        omitted: number[];
        p: number;
        press: number;
        r2: number;
        r2adj: number;
        r2pred: number;
        residuals: Float64Array;
        s: number;
        standardizedResiduals: Float64Array;
        studentizedResiduals: Float64Array;
        terms: TermRow[];
        test: "regression";
        unusual: {
            fitted: number;
            flags: string;
            index: number;
            leverage: number;
            residual: number;
            stdResidual: number;
            y: number;
        }[];
        y: Float64Array;
        predict(
            x: ArrayLike<number> | ArrayLike<number>[],
            options?: { confidence?: number },
        ): Prediction[];
    }
    Index
    aic aicc anova bic coefficients confidence cooksD covariance design dffits durbinWatson fitted intercept leverage logLik n names omitted p press r2 r2adj r2pred residuals s standardizedResiduals studentizedResiduals terms test unusual y
    predict
    aic: number
    aicc: number
    anova: { error: AnovaRow; regression: AnovaRow; total: AnovaRow }
    bic: number
    coefficients: Coefficient[]
    confidence: number
    cooksD: Float64Array
    covariance: Matrix

    Coefficient covariance matrix (p × p, in coefficient order).

    design: Matrix

    Design matrix used for the fit (after dropping missing rows).

    dffits: Float64Array
    durbinWatson: number
    fitted: Float64Array
    intercept: boolean
    leverage: Float64Array
    logLik: number
    n: number
    names: string[]
    omitted: number[]

    Row indices (of the input) dropped for missing values.

    p: number

    Number of estimated (non-aliased) coefficients including the constant.

    press: number
    r2: number
    r2adj: number
    r2pred: number
    residuals: Float64Array
    s: number
    standardizedResiduals: Float64Array
    studentizedResiduals: Float64Array

    Externally studentized (deleted-t) residuals.

    terms: TermRow[]
    test: "regression"
    unusual: {
        fitted: number;
        flags: string;
        index: number;
        leverage: number;
        residual: number;
        stdResidual: number;
        y: number;
    }[]

    Minitab's unusual observations: R = large standardized residual (|r| > 2), X = high leverage (h > 3p/n).

    y: Float64Array

    • Predict at new rows (arrays of predictor values in names order, without the constant).

      Parameters

      • x: ArrayLike<number> | ArrayLike<number>[]
      • Optionaloptions: { confidence?: number }

      Returns Prediction[]

    Settings
    Member Visibility

    On This Page
    Properties
    Methods