generate same number to one of the group

[rohithverma]

have input dataset  like 

DATA K;
INPUT city:$ ;
DATALINES;
a
a
a
a
a
b
b
b
;
RUN;

i need output like 

DATA r;

set k;

city seq;
a 1
a 1
a 1
a 1
a 1
b 2
b 2
b 2
;
RUN;

[PeterClemmensen]

If your data is sorted by city, then simply do

data r;
    set k;
    by city;
    if first.city then seq+1;
run;

If not then do

data r;
  if _n_=1 then do;
    dcl hash h();
    h.defineKey ("city");
    h.defineData ("seq");
    h.defineDone ();
  end;
 
  set K;

  if h.find() ne 0 then do;
     seq+1;h.add();
  end;
run;