*
* Autocorrelations of random walk and its differences, pp 342-34
*
all 300
set(first=1.0) y = y{1}+%ran(1.0)
*
* The BJIDENT procedure (which has been used before) also includes a DIFFS option
* to indicate the maximum number of differences to examine. With DIFFS=1, it will
* give (separate) graphs of correlations for both the levels (undifferenced) and
* first differenced series..
*
@bjident(diffs=1,number=12) y