Allin,

 

Thanks for the response.  Yes, you're correct-- basically, it entails applying the same filter 3 times to get a good smoothing of the data.

 

One other question: in the # of observations prompt, is this the value of "q"?

 

Thanks!

 

On 1/25/07, Allin Cottrell <cottrell@wfu.edu> wrote:

On Thu, 25 Jan 2007, Peter N. Krembs wrote:

> I'm still new to time-series analysis, and I have a question.
> I am confused about what process to use in order to create an
> iterated, moving-average filter for my time series data.  It
> needs to be smoothed by 3 iterations of a filter with a width of
> 30 days (my times series is monthly).  Should I use the simple
> moving average function for this?

I'm not familiar with the idea of an "iterated moving average" and
I can't find much reference to this in the modern econometric
literature.  But if it means what I guess it means, you could
create it by (a) using gretl's simple moving average filter to
create a new series, then (b) recursively applying the simple
moving average filter to the previous result, as many times as you
like.

Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users@ricardo.ecn.wfu.edu
http://ricardo.ecn.wfu.edu/mailman/listinfo/gretl-users