[S] ar.yw problem solved by Mathsoft

Bruce McCullough (BMCCULLO@fcc.gov)
Tue, 07 Apr 1998 16:10:13 -0500

I received a quick response from Mathsoft
on ar.yw, and we are getting to the bottom
of this ar.burg stuff. Text from Mathsoft follows:

I think that everything is actually
ok here (or at least I can get S-Plus to give
the same numbers as you are getting in RATS
and TSP). The discrepancy involves the
distinction between the autoregressive coefficients
and partial autocorrelation coefficients. What
you got when you typed aryw$ar was the former;
if you type

> aryw$partialacf

, , 1
[1,] 0.806243896
[2,] -0.634121358
[3,] 0.080473751
[4,] -0.061136235
[5,] 0.001139772

you get the partial autocorrelation coefficients.
Note that,when rounded to 2 digits, these numbers
are the same as the ones you reported that RATS & TSP
are giving you -- I suspect this might be the source of the

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