The easiest would be to add the year in the select and the group by statements: proc sql ;   select applprovenance, YEAR format=whatever sas date format you want,   sum(monoperl) as monoperl label= 'MontantOperation <10000',   sum(monoperg) as monoperg label= 'MontantOperation >=10000',   sum(monbilll) as monbilll label= 'MontantBilletEspece <10000',   sum(monbillg) as monbillg label= 'MontantBilletEspece >=10000'   from have2   group by applprovenance, year   ; quit; This doesnt create a new variable but adds the year column to the report. If you need the applprovenance and year in the same variable you could combine the two in a new varaible in the datastep: newvar = catx('_',applprovenance,year); You would then replace applprovenance (two times) with the new variable in SQL. Hope this helps! EJ ... View more

You can always recode the variables and use SQL to produce the report: data have2;   set have;   monoperl=0; monoperg=0; monbilll=0; monbillg=0;   if MontantOperation < 10000 then monoperl = 1;   else monoperg = 1;    if MontantBilletEspece < 10000 then monbilll = 1;   else monbillg = 1; run; proc sql ;   select applprovenance,   sum(monoperl) as monoperl label= 'MontantOperation <10000',   sum(monoperg) as monoperg label= 'MontantOperation >=10000',   sum(monbilll) as monbilll label= 'MontantBilletEspece <10000',   sum(monbillg) as monbillg label= 'MontantBilletEspece >=10000'   from have2   group by applprovenance   ; quit; ... View more

I think we are all getting to the same point ... but here is my version: data test3;   set test2;   if n = 1 then cat = 0;   group = mod(n,5);   if group = 1 then cat + 1; run; the group variable is basically providing an indicator for every 5th observation. The cat variable is categorizing the group of 5 observations (1-5 cat 1 , 6 - 10 cat2 , etc...). I wasnt clear if you are average the five observations (average obs 1 through 5, 6 through 10, etc... which would be over 8000 averages) or eliminate all but every 5th obs and average those obs for an outcome of one number. Anyway .. hope this helps! EJ ... View more

Can you provide more details about what you are trying to do and what you have tried? Without context it will be hard to help. ... View more

Is there a problem doing Table (A B C D E)*(A B C D E)?  There will be 5 extra tables (AA BB CC DD EE) but those could be deleted out using ODS document features. Just a thought, EJ ... View more