|
hello jim
here is one way... using a merge. i can think of another using a macro
approach depending on how you the end result should look.
HTH
Magnus
/* ---------- code... briefly tested on 8.2 ------------------- */
/* -- You original data sets -- */
data work.dataset1;
input year 1-4 nb 8-12;
cards;
1970 195
1980 205
;
run;
data work.dataset2;
input year 1-4 score 8-12;
cards;
1970 20
1970 30
1970 35
1980 10
1980 15
1980 06
;
run;
/* -- one delete list coming up -- */
data work.deletelist(keep=year delete_me);
set work.dataset1;
if (nb > 200); /* Set delete flag based on your nb > 200 */
delete_me = 1;
run;
/* add on the delete flags to your data and delete */
proc sql noprint;
create table work.result as
select a.*,
b.delete_me
from work.dataset2 a left join work.deletelist b
on (a.year=b.year)
where (delete_me ne 1); /* <<<<---- delete your unwanted ---
*/
quit;
/* ---------- end of code... briefly tested on 8.2 ------------------- */
On Mon, 15 Jul 2002 19:22:23 -0400, tin-shun-jimmy chan <jimmy.chan@HEC.CA>
wrote:
>Hello,
>
>I want to know how to do something in one dataset according to some
>conditions in another dataset.
>
>For example, if nb > 200 for the i-th year in one dataset, then delete
>all the observations of the whole i-th year in another dataset.
>
>dataset 1
>
>year nb
>1970 195
>1980 205
>...
>
>dataset 2
>year score
>1970 20
>1970 30
>1970 35
>1980 10
>1980 15
>1980 06
>...
>
>So I want to delete all the 1980's observations in dataset 2 according
>to dataset 1.
>
>Thanks a lot.
>
>Jim
| 
