Recode using ranges in data step?

Reply
Frequent Contributor
Posts: 75

Recode using ranges in data step?

[ Edited ]

I'm trying to recode age into groups but this is not working (just takes the first number in the bracket, from what I can tell)

 

data new;

set have;

 

if age in (15-24) then  agegroup=1;

else if age in (25-54) then agegroup=2;

else if age >= 55 then agegroup=3;

 

run;

 

What is the proper way to do this?

 

 

Edit: Also I know one can do it the way below, but it is very tedious, especially when there are more than 3 values being made. I'm looking for something easier, with ranges:

 

IF       15      <= age_tabs <=  24      then     agegroup_main = 1;      
else if  25      <= age_tabs <=  54      then     agegroup_main =2;      
else if             age_tabs >=   55     then     agegroup_main =3;      
else                agegroup_main = .
Grand Advisor
Posts: 10,241

Re: Recode using ranges in data step?

One way is to not add any variables an use your existing data set and assign a format to the variable.

 

proc format;

value agegrp

1 - 24 = '1'

25 - 54= '2'

55 - high='3';

run;

 

proc freq data=have;

   tables age;

   format age agegrp.;

run;

 

The format will create the group for the counts as well as changing displayed value.

 

Your code is not correct, use the colon : to indicate a numeric range with the IN operator

if age in (15:24) then  agegroup=1;

else if age in (25:54) then agegroup=2;

else if age >= 55 then agegroup=3;

 

Ask a Question
Discussion stats
  • 1 reply
  • 113 views
  • 2 likes
  • 2 in conversation