*
*  Example 12.4 from page 405
*
cal(m) 1978:2
open data barium.raw
data(org=columns) 1978:2 1988:12 chnimp bchlimp befile6 affile6 afdec6 befile12 affile12 afdec12 $
  chempi gas rtwex spr sum fall lchnimp lgas lrtwex lchempi trend feb mar apr may jun $
  jul aug sep oct nov dec percchn
*
* Least squares
*
linreg lchnimp
# constant lchempi lgas lrtwex befile6 affile6 afdec6
*
* AR1 (Cochrane-Orcutt). The default method for AR1 in RATS is called
* Hildreth-Lu, which is a grid search, rather than an iterative process. This
* should produce the same results as Cochrane-Orcutt. If it doesn't,
* Cochrane-Orcutt hit a local minimum of the sum of squares.
*
* The results are very slightly different from those in the text probably because
* of slight differences in the estimate of rho due to different convergence
* criteria.
*
* You'll notice that the RATS R**2 is higher in the AR1 case, while it's listed
* as lower in the text. RATS is using the original dependent variable in
* computing the R**2, not the transformed one, so the R**2's are comparable
* between the AR1 and LINREG.
*
ar1(method=corc) lchnimp
# constant lchempi lgas lrtwex befile6 affile6 afdec6
*
* Using REPORT to generate most of the table on page 406
*
report(action=define,hlabels=||"Coefficient","OLS","Cochrane-Orcutt"||)
linreg lchnimp
# constant lchempi lgas lrtwex befile6 affile6 afdec6
report(regressors)
ar1(method=corc) lchnimp
# constant lchempi lgas lrtwex befile6 affile6 afdec6
report(regressors)
report(row=new,col=current) %rho
*
* This formats all the numbers with three digits right of the decimal With a
* series of REPORT(ACTION=FORMAT) instructions using the ATROW and TOROW options,
* it would be possible to get the formatting to match the pattern given in the
* text.
*
report(action=format,picture="*.###")
report(action=show)