transposing to make one record per id.

[vraj1]

I am stuck up with transposing as i get multople observations if i try to transpose as the paramcd has multiple observations. attached code i am using and i need single observation per subject. any suggestion where am i going wrong

proc sort data=Anl1
	           out= anl2;
     by studyid usubjid paramcd aval;
run;

data Id_adds (keep=studyid usubjid aval avalc paramcd seq);
      set anl2;
      by studyid usubjid paramcd aval;
      seq+1;
       if first.paramcd then seq=1;
run;

proc transpose data=Id_adds out=Id_trans(drop=_name_ _label_) ;
      by studyid usubjid;
     id paramcd seq;
     var aval;
run;

[RW9 · Accepted solution]

This:

     id paramcd seq;

Wont work.   Update here:

data Id_adds (keep=studyid usubjid aval avalc paramcd seq);
  set anl2;
  by studyid usubjid paramcd aval;
  retain seq;
  if first.paramcd then seq=1; 
  else seq+1;
  paramcd=cats(paramcd,put(seq,best.));
run;

Then put:

proc transpose data=Id_adds out=Id_trans(drop=_name_ _label_) ;
  by studyid usubjid;
  id paramcd;
  var aval;
run;

Ie make the variable paramcd cd unique by adding the incremental number to it (when you have the sequence correct).