Getting second lowest value By group

[nstdt]

I know that the following SQL code is used to get the second-lowest value overall from a table:

PROC SQL;
SELECT id, min (value) 
FROM table_name
WHERE value NOT IN (SELECT min (value) 
                          FROM table_name); 

How can I modify the Sql code I have to achieve this? Or is there any other way to do this?

[PaigeMiller]

Don't use SQL. Use PROC SORT or PROC RANK.

[yabwon · Accepted solution]

data step it for this task in SAS:

data have;
input id value;
cards;
1 1
1 2 
1 3
3 300
3 200
3 100
2 3
2 4 
2 5
;
run;
proc print data=have;
run;


proc sort data=have out=want;
  by id value;
run;

data want;
  set want;
  by id;

  if first.id then Marker=0;
  Marker+1;

  if marker = 2;
  drop Marker;
run;

proc print data=want;
run;

Bart