sum of delhi and chandigarh only

[aanan1417]

data have;

input state $ income;

datalines;

delhi             1000

delhi             3000

chandigarh   4000

chandigarh   3000

patna             300

delhi               400

;

run;

data want;

input state $ income;

datalines;

delhi             4400

chandigarh   7000

;

run;

[Kurt_Bremser]

Use a where= dataset option:

proc summary data=have (where=(state in ("...","...")));
by state;
var income;
run;

[SASJedi · Accepted solution]

As written, the length of the state variable in your have and want datasets will default to 8 - too small to hold all the values. Try this:

data have;
length state $ 10;
input state $ income;
datalines;
delhi             1000
delhi             3000
chandigarh   4000
chandigarh   3000
patna             300
delhi               400
;
run;

 

data want;
length state $ 10;
input state $ income;
datalines;
chandigarh   7000
delhi             4400
;

run;

proc sql;
/*create table want as */
select state, sum(income)
   from have
   where lowcase(state) in ('delhi','chandigarh')
   group by state
;
quit;

May the SAS be with you 😉