Is there was a combine the two sql steps below into one?
/*Sample data*/
data one;
input x $1. y 1.;
datalines;
a1
a2
a1
b1
b1
;
run;
proc sql;
/*determine count of distinct values of y for each value of x*/
create table two as select x, count(distinct y) as cnt_y
from one
group by x;
/*determine maximum value of distinct value count*/
select max(cnt_y)
from two;
quit;
Hi @Batman You could do in one step that still doesn't mean it's a one pass though. For brevity, you could modify the code to-
proc sql;
select max(cnt_y)
from (select count(distinct y) as cnt_y from one group by x);
quit;
However the above is still 2 passes akin to your solution
Hi @Batman You could do in one step that still doesn't mean it's a one pass though. For brevity, you could modify the code to-
proc sql;
select max(cnt_y)
from (select count(distinct y) as cnt_y from one group by x);
quit;
However the above is still 2 passes akin to your solution
April 27 – 30 | Gaylord Texan | Grapevine, Texas
Walk in ready to learn. Walk out ready to deliver. This is the data and AI conference you can't afford to miss.
Register now and save with the early bird rate—just $795!
Learn how use the CAT functions in SAS to join values from multiple variables into a single value.
Find more tutorials on the SAS Users YouTube channel.
Ready to level-up your skills? Choose your own adventure.