s-news
[Top] [All Lists]

Summary: running means & running medians

To: s-news@lists.biostat.wustl.edu
Subject: Summary: running means & running medians
From: FolkesM@pac.dfo-mpo.gc.ca
Date: Wed, 21 Nov 2001 19:18:36 -0500
Wee Haa!
I hit a nerve there.
My inbox was awash with responses. Was it because it was easy or an issue?
I knew it would be an issue (sensitive to anomalous values) but was keen to
see the obvious.
I won't cover all answers or I'll be here for the evening.
The bottom (easiest) line is use filter(), running.ave seems to be
temperamental, I assume it had something to do with NA's but didn't pursue
it...

David Smith:
"Running means is easy to do with the function "filter".  For example:
filter(x,rep(1,3)/3)
does running means with a window size of 3."


>In looking across my options for presenting smoothed data I can't
>find anywhere a function to calculate running average (of varying 'window'
sizes).
>  I've used the various lowess and smooth functions, but was wondering why 
>running averages are apparently completely absent from Splus, 
> Tukey (1977), and both of Cleveland's books?  Is the running average
unacceptable in the eyes of some?


Thanks
Michael
________________________________________________________
Michael Folkes                  folkesm@pac.dfo-mpo.gc.ca
Fisheries & Oceans Canada               Tel (250) 756-7264
South Coast                             Fax (250) 756-7162
3225 Stephenson Point Road
Nanaimo, B.C.
Canada V9T 1K3 


<Prev in Thread] Current Thread [Next in Thread>
  • Summary: running means & running medians, FolkesM <=