我的意思是你的数据"y",
处理过程有没笔误,因为看不出有Business Cycle.
底下我提供Terasvirta JASA 1994
Terasvirta(1994), "Specification, Estimation and Evaluation of
Smooth Transition Autoregressive Models", JASA, vol 89,
data "lynx",及 "ipgrowth"的图型供你参考.
#####
library(tsDyn)
#data lynx
#data ipgrowth
ip =read.table("ipgrowth.txt", header = TRUE);
ipgrowth =ip$ipgrowth
#your data y
new=read.table("x11.txt",header=TRUE)
y=new$y
#plot
par(mfrow = c(2, 2))
plot(log10(lynx))
lines(log10(lynx),col='red')
plot(ipgrowth)
lines(ipgrowth,col='red')
plot(y)
lines(y,col='red')