Hi, I encountered some problem in finding seasonal component. I wish I could find some suggestions.The thing is for a time series, assume I have no idea about where this serie comes from but I still want to know whether it has seasonal component or not, then what should I do? Also assume I know the PACF and ACF.

Whats the dimension of your data?

It's a time series observations against the time index. Any suggestions are welcome

periodogram?

why dont you just run a standard OLS on the series against time, plot the residuals / ACF of residuals, and see if you get a pattern?

a sound advice, thx.Any further suggestion on finding seasonal component based on formula? Or is there some reference book?

Have you looked at Harvey's papers?

have a look at Davidson and MacKinnon "Econometric theory and Methods". Excellent reference for any econometrics question.also Hamilton in his "Time Series Analysis" deals with that. if you know or can think of the dgp for the data, then throwing in the dummy variables for the seasons will give you an idea of the significance of the seasons through the coefficient on the season variable. i am not sure periodogram will help. it just tells you the changes in the variable that are attributable to different frequencies, i do not see how you would infer seasonality from there. perhaps Rez could clarify what I am missing.

Hi there,The periodogram has the advantage of being model-free, and you don't need to specify a DGP. Given seasonality, the periodogram should 'spike' identifying the cycles, which would be the reciprocal of the seasonality periods. Multiple spikes would identify shorter and longer seasonal components (e.g. there might be a yearly and a weekly seasonal pattern). For a non-technical overview you can check out this SPSS manual, chapter 13.Cheers

I think the Winter method is very useful for seasonalitry anslysis.Any standard stats text will contain the details and also be found on the internet.

Thanks for all the recommendations, I'd test them one by one THe chapter 6 Spectral analysis of Hamilton is a good introdution

Thanks for all the recommendations, I'd test them one by one THe chapter 6 Spectral analysis of Hamilton is a good introdution

Thanks for all the recommendations, I'd test them one by one THe chapter 6 Spectral analysis of Hamilton is a good introdution

Thanks for all the recommendations, I'd test them one by one THe chapter 6 Spectral analysis of Hamilton is a good introdution

Thanks for all the recommendations, I'd test them one by one THe chapter 6 Spectral analysis of Hamilton is a good introdution