Carry out Tsay's test for quadratic nonlinearity in a time series.
Usage
Tsay.test(x, order, ...)
Arguments
x
time series
order
working linear AR order; if missing, it will be estimated
via the ar function by minimizing AIC
...
options to be passed to the ar function
Value
A list containing the following components
test.statThe observed test statistic
p.valuep-value of the test
orderworking AR order
Details
The null hypothesis is that the true model is an AR process. The AR order, if missing, is estimated by minimizing AIC via the
ar function, i.e. fitting autoregressive model to the data. The default
fitting method of the ar function is "yule-walker."
References
Tsay, R. S. (1986), Nonlinearity test for time series,
Biometrika, 73, 461-466.