*
* Example 5.2.4
*
open data dowj.dat
data(format=free,org=columns) 1 78 dowj
*
set ddow = dowj-dowj{1}
*
* This estimates the model by maximum likelihood, after extracting the sample
* mean.
*
boxjenk(ar=1,demean,maxl) ddow
*
* This will produce slightly different results by estimating the mean as part of
* the model. (The estimates of the mean differ because the error variance for the
* full model is slightly higher for the first few data points, so they won't get
* equal weight in estimating the mean).
*
boxjenk(ar=1,constant,maxl) ddow
*
* This is an alternative to the original set of estimates. Instead of
* transforming the series to differences separately, you can include a DIFFS
* option on the BOXJENK and apply the instruction to the original series. Note
* that the "de-meaning" takes place after the data have been differenced
*
boxjenk(ar=1,diffs=1,demean,maxl) dowj