*
* Box-Jenkins identifications for consumption/income data
* pp 361-366
*
open data consump.dat
calendar 1950
data(format=prn,org=columns) 1950:01 1993:01 year y c
*
graph(footer="Figure 14.1 U.S. Consumption and Income 1950-1993",$
 klabels=||"Consumption","Income"||,key=attached) 2
# c
# y
*
set logc = log(c)
set logy = log(y)
*
* The METHOD=YULE option chooses the same algorithm as is used in the table in
* the book. The default is METHOD=BURG, which is considered to provide superior
* estimates for a non-stationary or near non-stationary series like this one.
*
@bjident(report,number=13,method=yule) logc