KPSS.N.breaks.bootstrap: KPSS-test with multiple unknown structural breaks

In d9d6ka/RANEPA-R: Set of unit root and cointegration tests

KPSS-test with multiple unknown structural breaks

Description

Procedure to compute the KPSS test with multiple unknown structural breaks

Usage

KPSS.N.breaks.bootstrap(
  y,
  x,
  model,
  break.point,
  const = FALSE,
  trend = FALSE,
  weakly.exog = TRUE,
  lags.init,
  leads.init,
  max.lag,
  kernel,
  iter = 9999,
  bootstrap = "sample",
  criterion = "bic"
)

Arguments

y	A time series of interest.
x	A matrix of explanatory stochastic regressors.
model	A scalar or vector of 1: for the break in const, 2: for the break in trend, 3: for the break in const and trend.
break.point	Array of structural breaks.
const	Include constant if TRUE.
trend	Include trend if TRUE.
weakly.exog	Boolean where we specify whether the stochastic regressors are exogenous or not TRUE: if the regressors are weakly exogenous, FALSE: if the regressors are not weakly exogenous (DOLS is used in this case).
lags.init, leads.init	Scalars defininig the initial number of lags and leads for DOLS.
max.lag	scalar, with the maximum order of the parametric correction. The final order of the parametric correction is selected using the BIC information criterion.
kernel	Kernel for calculating long-run variance bartlett: for Bartlett kernel, quadratic: for Quadratic Spectral kernel, NULL for the Kurozumi's proposal, using Bartlett kernel.
iter	Number of bootstrap iterations.
bootstrap	Type of bootstrapping: "sample": sampling from residuals with replacement, "Cavaliere-Taylor": multiplying residuals by N(0, 1)-distributed variable, "Rademacher": multiplying residuals by Rademacher-distributed variable.
criterion	Information criterion for DOLS lags and leads selection: aic, bic or lwz.

Value

A list of:

• test: The value of KPSS test statistic,

• p.value: The estimates p-value,

• bootstrapped: Bootstrapped auxiliary statistics.

d9d6ka/RANEPA-R documentation built on May 4, 2024, 7:11 a.m.