create 2 text files, append them and then sort them

[DanD999]

Each file would have a different number of columns. We would want to sort by the id column. Is this possible in 9.4?

data file1;
input @1 id 1. @3 start_date mmddyy10.;
format adm_date mmddyy10.;
cards;
1 03/18/2014 
1 03/27/2014
1 03/24/2014
2 03/19/2014
3 03/24/2014
3 03/25/2014

;
run;


data file2;
input @1 id 1. @3 proc_code $5. @9 app_units 1. @11 request_units 1.;
/*format start_date end_date mmddyy10.;*/
cards;
1 J5563 2 4
1 J5563 3 1
2 J3363 4 2
2 J4789 2 6
3 J8434 2 2 
3 J4444 5 5 
;
run;

1 03/18/2014 
1 03/27/2014
1 03/24/2014
1 J5563 2 4
1 J5563 3 1
2 03/19/2014
2 J3363 4 2
2 J4789 2 6
3 03/24/2014
3 03/25/2014
3 J8434 2 2
3 J4444 5 5

[ballardw]

Your example output tends to imply that you have combined the two datasets and are exporting to one file, not two.

This creates a data set with contents that look like your output. I have removed the blank line generated in your data step for file1.

The OPTIONS MISSING=' '; is so that the bit that combines all of the variables in the temporary set don't get . The catx places a single space between all of the non-missing values that way. If your real data has more variables you may need to increase the length of the Line variable to accomodate more or longer variables.

data file1;
input @1 id 1. @3 start_date mmddyy10.;
format adm_date mmddyy10.;
cards;
1 03/18/2014 
1 03/27/2014
1 03/24/2014
2 03/19/2014
3 03/24/2014
3 03/25/2014
;
run;


data file2;
input @1 id 1. @3 proc_code $5. @9 app_units 1. @11 request_units 1.;
/*format start_date end_date mmddyy10.;*/
cards;
1 J5563 2 4
1 J5563 3 1
2 J3363 4 2
2 J4789 2 6
3 J8434 2 2 
3 J4444 5 5 
;
run;

options missing=' ';
data temp;
   set file1 file2;
   length line $ 200;
   line= catx(' ',of _all_);
run;
proc sort data=temp out=want (keep=line);
   by id;
run;

proc export data=want out='path/file.txt';
run;

options missing='.';

[DanD999]

Thanks so much for the reply. It looks like a winner.

[DanD999]

After looking at this closer, I don't think it works. Here is the output.

1 19800
1 19809
1 19806
1 J5563 2 4
1 J5563 3 1
2 19801
2 J3363 4 2
2 J4789 2 6
3 19806
3 19807

It combines the start_date and proc_code columns into one column and they should be two separate columns.
3 J8434 2 2
3 J4444 5 5

[DanD999]

I took out the options missing=' '; line and it looks like that may work. It preserves each column. I'll have to look at the actual data but I think this might work. Thanks for the help.