csv file with column names

[GreggB]

I need to amend this data step so that the column names are written to output file in row 1.

filename csv "G:\Departments\Research\MAP\&yr1&yr2.\Fall &yr.\clean_for_enrich.csv";

data _null_;
set join1b ;
file csv dlm=',';
put ( _all_ ) (+0);
run;

[ballardw · Accepted solution]

Easiest may be

Proc export data=join1b outfile=csv dbms=csv;

run;

[GreggB]

Agreed. I was making it  too complicated.

[ballardw]

If you need to provide a specific order, such for a file that may be read by another program that requires a fixed order then the PUT statement would be appropriate but since you didn't show that then I though export would work.

The easy semi-kludge for the header row is to include something like this BEFORE the Put statement in the order to match the variables on the put statement:

If _n_ = 1 then put '"column 1 text","column 2 text", ...,"column n text"';

note that is '  "   and "  ' at the start and end. If there are no spaces or commas desired in the Column x text then the " could be omitted.

The line basically says: If this is the first line of the file then put the header row.