pcdiff.piar2unit.block: Periodically differenced data when there are two block unit...

In YueyunZhu/UnitRootsofPIAR: Eliminate unit roots for periodically integrated autoregressive models

Description

Periodically differenced data when there are two block unit roots in original data

Usage

pcdiff.piar2unit.block(pcts, parameters, d = 4)

Arguments

pcts	the original time series (assumed there are two block unit roots)
parameters	parameters of second order differencing filter $(1-\beta_sL)(1-\alpha_sL)$
d	period has been fixed to 4 (quarterly period)

Value

Periodically differenced data

Examples

library(pcts)
library(mcompanion)
set.seed(123)
F <- sim_mc(dim = 4, mo = 4, mo.col = 2, eigval = c(1), len.block = c(2))
phi <- new("MultiFilter", mc =  mCompanion(F$mat, mo = 4, mo.col = 2))
pc.sim <- sim_pc(model = list(phi = phi[], p = 2, period = 4), n = 500)
# coefficients of second order filter
icoef <- piar2unit.block(Xsim = F$eigvec, d = 4)
pcdiff.pc.sim <- pcdiff.piar2unit.block(pcts = pc.sim, parameters = icoef, d=4)

YueyunZhu/UnitRootsofPIAR documentation built on Feb. 13, 2020, 12:32 a.m.