Hi, I have a question about the ar function in the stats package, it is a method to use autoregressive models for time series. Now I have a time series, which I performed a spectral analysis on. This gives a spectrum with a quite impressive peak at a certain frequency. The AR1 function I want to use to statistically test if this peak is significant or if it is produced by so called 'red noise'. For me it is not very clear how the ar() function works, even with studying the chapter in the R documentation. - Should I for example apply the function on the original series or on the spectral analysis data. If I do the last, then how does R know which frequencies belong to the spectral densities? - How can I use the ar coefficient to plot the line of 95% interval? Thank you for your help. [[alternative HTML version deleted]]