topic Re: create list of missing variables by subject in SAS Programming

create list of missing variables by subject

aklr — Tue, 08 May 2012 15:09:07 GMT

Hello,

I have a dataset in which each line is a subject, and I'm interested in creating a new variable that contains a list of 40 possible variables, but only the ones that are missing for each subject. Is there an existing macro for this task? Can anyone help me figure out how to create a variable that contains a list of variables that is unique to each line in a dataset?

Re: create list of missing variables by subject

Linlin — Tue, 08 May 2012 15:12:34 GMT

are the variables numeric or character or both?

Re: create list of missing variables by subject

aklr — Tue, 08 May 2012 15:14:59 GMT

They are all numeric.

Re: create list of missing variables by subject

Linlin — Tue, 08 May 2012 15:23:46 GMT

try:

improved version:

data have;

input id $ v1 v2 v3 v4;

cards;

a 1 2 3 .

b 2 3 . .

c 2 . . .

;

data want;

length new $ 50;

set have;

array _v(*) _numeric_;

do _n_=1 to dim(_v);

if _v(_n_)=. then new=catx('-',new,vname(_v(_n_)));

end;

proc print;run;

Linlin

Message was edited by: Linlin

Re: create list of missing variables by subject

aklr — Tue, 08 May 2012 15:44:47 GMT

I needed to get rid of the retain. Otherwise I was getting repeated variable names in the list, but otherwise, this is perfect. Exactly what I needed. Thank you!

Re: create list of missing variables by subject

Linlin — Tue, 08 May 2012 16:02:12 GMT

Hi,

"call missing (your-variable)" gets rid of the missing variable names from previous IDs.

Re: create list of missing variables by subject

PGStats — Tue, 08 May 2012 16:08:29 GMT

Hi Linlin, since you create a new list for each observation, you don't need the retain, the call missing() or the output statements. It works just as well without them.

Re: create list of missing variables by subject

Linlin — Tue, 08 May 2012 16:13:18 GMT

Thank you PG!!!

improved version:

data have;

input id $ v1 v2 v3 v4;

cards;

a 1 2 3 .

b 2 3 . .

c 2 . . .

;

data want;

length new $ 50;

set have;

array _v(*) _numeric_;

do _n_=1 to dim(_v);

if _v(_n_)=. then new=catx('-',new,vname(_v(_n_)));

end;

proc print;run;