Hi @ebowen
It is quite difficult without the data but maybe you could try to start with something like this to adapt the code:
proc report data=lq3 out=want (drop=_break_);
by region;
columns clustercode clustername employment_R, year employment_US, year
Employ_tot_R, year Employ_tot_US, year income_R, year income_US, year Establish_R, year
Establish_US, year employment_R, pctsum employment_US, pctsum;
define clustercode/group '';
define clustername/group '';
define year / across '';
run; proc print data=want ; run;
- The page dimension of your PROC TABULATE (= by region) can be handled with a BY statement
- The row dimension can be handle with the GROUP option in define statements (cluster code, cluster name)
- The year can be define as an ACROSS variable to obtain as many column as years for each variable (stack under them).
- For analysis variables (numeric ones), the default statistic is SUM. For the last two ones, you can request PCTSUM by adding this statistic after a comma.
Best,
... View more
