Interface DlmForecastResult

Result from dlmForecast: h-step-ahead predictions with uncertainty. Uses StateMatrix and CovMatrix for predicted state trajectories.

interface DlmForecastResult {
    h: number;
    m: number;
    predicted: StateMatrix;
    predictedCov: CovMatrix;
    predictedStd: StateMatrix;
    yhat: FloatArray;
    ystd: FloatArray;
}

Properties

h: number

Forecast horizon

m: number

State dimension

Predicted state means [h, m].

predictedCov: CovMatrix

Predicted state covariances [h, m, m].

predictedStd: StateMatrix

Predicted state std devs [h, m] = sqrt(diag(predictedCov)).

Predicted observation means (length h). yhat[k] = F · predicted state at step k+1.

Predicted observation std devs (length h). ystd[k] = sqrt(F·C·F' + s²). Monotonically increasing.