Unit Root Test
Unit root test. Test H0: rho=1 vs. H1: rho<1 in the model model with intercept z[t] = const + rho*z[t-1] + a[t].
UnitRootTest(z, method = c("MLE", "ExactMLE", "LS", "All"), statistic = c("Z", "T"), NumBoot = 1000, PValueMethod = c("DH", "ET"))
z |
time series |
method |
estimation methods |
statistic |
normalized rho or t-statistic |
NumBoot |
number of bootstrap iterations |
PValueMethod |
p-value can be estimated either as (k+1)/(N+1) as recommended by Davison and Hinkley (p. 148) or as k/N as in Efron and Tibsharini (p. 221, Algorithm 16.1 ). |
Bootstrap unit root tests
one-sided P-value
A.I. McLeod
Davison, A.C. and Hinkley, D.V. (1997). Bootstrap Methods and their Application. Cambridge.
Efron, B. and Tibshirani, R. (1993). An Introduction to the Bootstrap. Chapman/Hall.
Yu, H., Zhang, Y. and McLeod, A.I. (2009). Unit Root Bootstrap Tests with Exact Maximum Likelihood.
## Not run: #takes about 10 seconds z<-cumsum(rnorm(100)) UnitRootTest(z) ## End(Not run)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.