[S] serial correlation

Charles Pollak (Charles_Pollak@intmed.med.ohio-state.edu)
Sat, 27 Jun 98 20:04:32 EST


Dear all:

I want to generate 2 parallet time series that differ in the
degree of serial correlation. I tried:

a <- rexp(100, rate = 1/250)

x <- arima.sim(innov=a, model=list(ar=.7))

but mean(x) >> mean(a) ... typically 600 vs. 250 for the given
parameters. Can anybody tell me why the means differ? How can
I manipulate the degree of serial correlation without strongly
affecting the means? Thanks in advance.

-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news