[R] Time series exercise (was: (no subject)).

[Rolf Turner]

On 10/07/15 21:17, Lia LEE wrote:
> Hello forum members,
>
> I am taking this R course which I have to admit that it is beyond my
> capability. I am asked to analyze the following data(file attached) according
> to the question provided below:
>
> 1. The file [data_13-9.txt] contains a data of length 225, which seems to
> have some cycle pattern.
> (1) Fit a suitable model for this data, then report the model equation.
> (2) Provide the forecast values for 3-steps ahead and 6-steps ahead.
>
> I am desperately seeking for your help, since I really don't know where to
> start.. I have done "plot.ts(data.13-6)", but not further. Above question
> is just part of the whole task, but I believe solving this would be very
> much helpful for me to jump in to the other questions. It would be easier
> for me to follow if your answer is in R script..
> I appreciate your patience and help.
> Thank you so much in advance!

As Bert Gunter said, we don't do homework for people on this list. 
Notwithstanding I couldn't resist having a crack at analysing your time 
series, and I found it, uh, challenging to say the least.  I think of 
myself as having at least *some* skill at fitting time series models, 
and I struggled to find a model that fitted adequately to your data.
(I finally managed to get a decent fit, after trying a number of models 
on a hammer-and-hope basis.)

Perhaps I am even more of a thicko than I think I am.  Nevertheless it 
seems to me that your instructor has dropped you in the deep end by 
setting you this exercise.  (Perhaps I am missing something obvious, as 
is so often the case.  Others may wish to chime in with some comments to 
the effect "You idiot Rolf.  The *obvious* model is ....")

At any rate, your first stop should be your instructor.  Get him or her 
to give you some pointers as to how to get started.  That is what 
instructors are for.

cheers,

Rolf Turner

Technical Editor ANZJS
Department of Statistics
University of Auckland
Phone: +64-9-373-7599 ext. 88276