
The time series is not stationary. So there is a slowly decaying ACF.
library(astsa)
acf(jj)
pacf(jj)
Differenced to remove trend.
\[ y_t=x_t-x_{t-1} \]
jj.d <- diff(log(jj))
Trend removed.
plot(jj.d)
Now we can see the seasonal part of the time series.
acf(jj.d)
Seasonally differenced to remove the seasonality.
\[ y_t=x_t-x_t-d \]
Where d is 4 because this is quarterly data.
jj.d <- diff(log(jj), 4)
This looks more random.
plot(jj.d)
No clear pattern.
acf(jj.d)