topic Re: How do i covert N and O to 1 and 0 ? in SAS Programming

How do i covert N and O to 1 and 0 ?

serakuru1989 — Mon, 24 Sep 2018 17:51:50 GMT

How do I covert N and O to 1 and 0 to do correlation and regression analysis? Please help me with this.

Re: How do i covert N and O to 1 and 0 ?

Reeza — Mon, 24 Sep 2018 17:54:28 GMT

What PROC are you using for regression?

If it's PROC LOGISTIC or GLM (and several others) you can use a CLASS statement to specify this is a categorical variable. You'll also want to specify the reference level but you don't need to convert them to 0/1 manually. Otherwise you can use IF/THEN statements to make a new variable.

@serakuru1989 wrote:

How do I covert N and O to 1 and 0 to do correlation and regression analysis? Please help me with this.

if varName = "N" then varName_Cat = 0;
else if varName = "O" then varName_Cat = 1;

Re: How do i covert N and O to 1 and 0 ?

serakuru1989 — Mon, 24 Sep 2018 17:59:33 GMT

I am very much new in this. Using PROC Logistic.

How do I write a CLASS statement to specify this is a categorical variable?

Re: How do i covert N and O to 1 and 0 ?

Reeza — Mon, 24 Sep 2018 18:01:45 GMT

Follow the documentation example then: https://documentation.sas.com/?docsetId=statug&docsetTarget=statug_logistic_examples02.htm&docsetVersion=14.3&locale=en

I would recommend running the documentation example first, make sure you understand it and then try it with your own data. The example doesn't do one thing I would recommend, set the class variables to have the REF parameterization.

class <list of categorical variables> / param=ref;