On Wed, Dec 22, 2010 at 10:09 AM, new_user
<address@hidden> wrote:
> some kind of fourier analysis to look for the periodicity?
Well, if my name suggests association... ;)
I'll try to open some books and reply with solution if I find it
(FYI, I'm just a meteorology student)
Autocorrelation ( if I remember correctly) makes a copy of your data and then tries shifting and overlapping your data with the copy. I then looks for a correlation between the 2 for each overlap and then reports this correlation, if you plot this correlation vector you will find a peek at the periodicity points