Recoding 2s, 3s, 4s as 1s for race

[klongway]

Hello again,

I've been banging my head on this and thought I'd turn to SAS Community for some advice.

I have a government data set. The data has 15 racial categories. It is NOT 0/1 coded like I originally thought. People can select up to 15 of the 15 categories (so, someone could choose from 1-15 identifiers).  The data is coded like 1=chinese/0-didn't choose, 2=native alaskan/0=didn't choose; 3= burmese/0=didn't choose.

I am struggling to code for identifying multiracial individuals. 

I had originally planned that 0=didn't choose and 1=chose that identifier, so anyone with a total of greater than 1 had chosen multiple racial identifiers and can be labeled "multiracial."

How would I change the non 1 numbers for "chose this" into 1s? So instead of a person choosing native alaskan (2) and burmese (3) and equaling 5, I would want them to equal 2 (native alaskan-1, burmese=1).

Hope this makes sense. Thanks!!!

[Tom · Accepted solution]

SAS treats 0 (or missing) as FALSE and any other number as TRUE.  So to convert your 2/0, 3/0,.... variables into 1/0 variables you can just do:

chinese = not not chinese;

So 0 will go to 1 and then back to 0.  Any other number, like 1,2,3,etc will go to 0 and then back to 1.

[klongway]

wow, i should have gotten that. Thanks!!!

[novinosrin]

data test;
 do chinese=0 to 10;
  chinese_test=^^chinese;
  output;
 end;
run;