KpssTest Class

public sealed class KpssTest : OneSampleTest<double>

Public NotInheritable Class KpssTest
	Inherits OneSampleTest(Of Double)

public ref class KpssTest sealed : public OneSampleTest<double>

[<SealedAttribute>]
type KpssTest = 
    class
        inherit OneSampleTest<float>
    end

Inheritance

Object  →  HypothesisTest  →  OneSampleTest<Double>  →  KpssTest

Use the KpssTest class to test whether a time series is level-stationary or trend-stationary. The null hypothesis is stationarity: stationarity around a constant for Level, or stationarity around a deterministic linear trend for Trend.

The Augmented Dickey-Fuller test has the opposite null hypothesis: a unit root. KPSS and ADF tests are often used together because the KPSS null hypothesis is stationarity while the ADF null hypothesis is nonstationarity due to a unit root.

Large KPSS statistics reject the null hypothesis. P-values are obtained by linear interpolation in the standard KPSS critical-value table and may be reported as bounds; see PValueStatus.

Reference