dlm-js
    Preparing search index...

    Interface DlmTensorResult

    On-device tensor result from dlmFitTensor. All arrays are np.Array tensors that stay on-device — no data transfer to JS. Implements Disposable for automatic cleanup via using.

    interface DlmTensorResult {
        deviance: Array;
        filtered: Array;
        filteredCov: Array;
        innovations: Array;
        innovationVar: Array;
        m: number;
        mape: Array;
        mse: Array;
        n: number;
        nobs: Array;
        rawResiduals: Array;
        residualVariance: Array;
        rss: Array;
        scaledResiduals: Array;
        smoothed: Array;
        smoothedCov: Array;
        smoothedStd: Array;
        standardizedResiduals: Array;
        yhat: Array;
        ystd: Array;
        "[dispose]"(): void;
    }

    Hierarchy

    • Disposable
      • DlmTensorResult

    Properties

    deviance: Array

    Deviance: -2 · log-likelihood (scalar)

    filtered: Array

    Filtered states [n, m]

    filteredCov: Array

    Filtered covariances [n, m, m]

    innovations: Array

    Innovations [n]

    innovationVar: Array

    Innovation variances [n]

    m: number

    State dimension

    mape: Array

    Mean absolute percentage error (scalar)

    mse: Array

    Mean squared error (scalar)

    n: number

    Number of observations

    nobs: Array

    Number of non-NaN observations (scalar)

    rawResiduals: Array

    Raw residuals [n]

    residualVariance: Array

    Residual variance (scalar)

    rss: Array

    Residual sum of squares (scalar)

    scaledResiduals: Array

    Scaled residuals [n]

    smoothed: Array

    Smoothed states [n, m]

    smoothedCov: Array

    Smoothed covariances [n, m, m]

    smoothedStd: Array

    Smoothed state standard deviations [n, m]

    standardizedResiduals: Array

    Standardized residuals [n]

    yhat: Array

    Fitted values [n]

    ystd: Array

    Prediction standard deviations [n]

    Methods

    Properties
    Methods