* Creating dicotomous variables.
* This works if there is no missing data.
compute hsplus = (educ >=12).
execute.

* Use this if there might be missing data.
if (~missing(educ)) hsplus2 = (educ >=12).
execute.

* "recode" is usually used with categorical input, but can
    also handle data ranges.
recode educ (8 thru 11 = 1 ) (12 thru highest = 2) into hsplus3.
* more typical use of recode.
recode jobcat (1 2 = 1) (3 4 = 2) into jobs2.
execute.

* Number cases within a group, here "gender"
    is a grouping variable.
sort cases by gender bdate.
compute seqnum = 1.
if (gender = lag(gender)) seqnum = lag(seqnum) + 1.
execute.

* Does not require the data to be sorted.
*  However, the by variable must be numeric, and
    it does not handle ties.
*sort cases by jobcat bdate.
rank variables=bdate by jobcat
  /rank into jobid.



sort cases by educ.
AGGREGATE
  /OUTFILE=* MODE=ADDVARIABLES
  /BREAK=educ
  /N_educ=N.
sort cases by n_educ.