Create a new variable by concatenating other non-missing variables

[TashaLChapman]

Hi folks,

Probably a common problem, and I'm just blanking on a solution that doesn't require 5,040 lines of code...

I have seven variables, of which any number of them can be populated or missing.  In any combination.  And I'd like to concatenate them into a single variables that displays the populated values as a list separated by commas.  But obviously I don't want extraneous commas, so I only want the new variable to include non-missing values.  Any thoughts on how to do this?

Original data:

Obs	Barrier1	Barrier2	Barrier3	Barrier4	Barrier5	Barrier6	Barrier7
1	Child	Funding		Title IV-E
2		Funding	Courts			Team
3					Policy	Team	Other
4	Child		Courts
5		Funding			Policy
6	Child		Courts	Title IV-E			Other

Desired new variable:

Obs	Barriers
1	Child, Funding, Title IV-E
2	Funding, Courts, Team
3	Policy, Team, Other
4	Child, Courts
5	Funding, Policy
6	Child, Courts, Title IV-E, Other

[Reeza · Accepted solution]

call catx function

[TashaLChapman]

Oh, duh.  It really is a Friday isn't it... where is my brain?  I'm so used to using the long-hand version with the concatenation operator I completely forgot about the catx function.  Thanks!

(That's 2 for 2 today, Reeza.)

[ballardw]

I would start with CATX function in a data step

data want;

     set have;

     length combinedBarriers $ <big enough integer to store longest combination>;

     combinedbarriers = catx(',' of Barrier1-Barrier7);

run;