Make some macro variable to hold it .
data have;
infile cards truncover expandtabs;
input ID Year Var1 Var2;
cards;
123 2012 345 24
310 2012 4 324
400 2012 234 456
123 2013 675 46
200 2013 577 89
310 2013 46 45
400 2013 45 76
123 2014 123 234
200 2014 1234 35
400 2014 435 645
123 2015 231 236
200 2015 456 87
310 2015 435 534
400 2015 232 235
123 2016 123 679
200 2016 354 523
310 2016 233 754
400 2016 200 342
;
run;
proc transpose data=have(obs=0 drop=id year) out=x;
run;
proc sql noprint;
select _name_ into : names separated by ',' from x;
select catt('array _',_name_,'{0:2} _temporary_;')
into : arrays separated by ' '
from x;
select catt('sum_',_name_,'=sum(of _',_name_,'{*});')
into : sum separated by ' '
from x;
select catt('_',_name_,'{mod(n,3)}=',_name_,';') into : v separated by ' '
from x;
quit;
proc sql;
create table temp as
select a.*,&names
from (select * from
(select distinct id from have),(select distinct year from have)) as a
left join have as b
on a.year=b.year and a.id=b.id
order by a.id,a.year;
quit;
data want;
set temp;
by id;
&arrays
if first.id then n=0;
n+1;
∑
&v
if n gt 3;
keep id year sum_:;
run;
