Here is another way. data have; input var1 var2 var3 var4 var5; id=_n_; datalines; 23    87    94    59   54 36    23    19    23   57 87    63    87    51   99 35    38    87    87   23 68    49    23    78   87 ; run; proc contents data = have   out = haveVars(keep = varnum name)   noprint; run; Proc Sql noprint;   Select count(name)   into :NmbrVrs separated by ''   from haveVars   where name contains "var"; Quit; %put &NmbrVrs   var&NmbrVrs; %macro fndDplcts();   Data lst (drop=var1 - var&NmbrVrs);   Set have;   %do p = 1 %to &NmbrVrs;   Nmbr=var&&p;   varNm="var&&p";   VarNmId=compress(varNm||"_Id"||_N_);   Cntr=1;   output;   %end;   Run; %mend; %fndDplcts; PROC SORT DATA=LST OUT=LstSrt;   BY Nmbr; RUN; PROC TRANSPOSE DATA=LstSrt OUT=LstTrnsps   NAME=Source LABEL=Label;   BY Nmbr;   ID VarNmId;   VAR Cntr; RUN; QUIT; Data LstTrnsps (drop= Source);   Set LstTrnsps;   array a(*) _numeric_;   do i=1 to dim(a);   if a(i) = . then a(i) = 0;   end;   drop i; run; proc contents data = LstTrnsps   out = LstTrnspsVars(keep = varnum name)   noprint; run; Proc Sql noprint;   Select name   into :VrsLst separated by ' '   from LstTrnspsVars   where name contains "var"; Quit; %put &VrsLst; Data LstCnt;   retain Nmbr Ttl &VrsLst;   Set LstTrnsps;   array vrArry &VrsLst;   Ttl=0;   do over vrArry;   Ttl=Ttl+vrArry;   end; run; Proc sort data=LstCnt;   by descending Ttl; run; ... View more