I have to identify which sneaker (variable 1) is matched (same row) with the highest and lowest sales (variable 2) by using a SAS procedure function.
Current code:
PROC MEANS DATA = M4.sneaker min max; Class salesThisPeriod; RUN;
You can try using the IDGROUP feature of the OUTPUT statement of PROC MEANS (aka PROC SUMMARY). But to get both min and max you will need to write two output datasets.
proc summary data=sashelp.class ;
class sex;
output out=min idgroup(min(age) out[1] (name age height weight)= );
output out=max idgroup(max(age) out[1] (name age height weight)= );
run;
data both;
length stat indsname $32 ;
set min max indsname=indsname;
by _type_ sex;
stat=scan(indsname,-1,'.');
run;
Result
Obs stat Sex _TYPE_ _FREQ_ Name Age Height Weight
1 MIN 0 19 Joyce 11 51.3 50.5
2 MAX 0 19 Philip 16 72.0 150.0
3 MIN F 1 9 Joyce 11 51.3 50.5
4 MAX F 1 9 Janet 15 62.5 112.5
5 MIN M 1 10 Thomas 11 57.5 85.0
6 MAX M 1 10 Philip 16 72.0 150.0