*
* Example from section 9.4
*
open data internet.dat
data(format=free,org=columns,skip=1) 1 100 internet
set diff = internet-internet{1}
*
* RATS has a procedure which creates a table such as the one shown in the text.
* We also show (3,0) as being the best model.
*
@bjautofit(pmax=5,qmax=5,crit=aic) diff
*
* This creates as "patch" the series with missing values inserted
*
set patch = diff
dofor drop = 6 16 26 36 46 56 66 72 73 74 75 76 86 96
   compute patch(drop)=%na
end dofor drop
*
@bjautofit(pmax=5,qmax=5,crit=aic) patch