@Julie4435637 @ChrisNZ @ChrisHemedinger : First, my apologies, my example was bad.
I may be mistaken but, I think that the order of concern (for nodup (noduprec) is dependent upon the order AFTER the sort .. not the one before the sort.
I was also wrong regarding that the difference was between character and numeric by variables.
The problem is simply that if a file has more than one variable, and a keep option is used to only include the by variable, nodup (noduprec) performs differently than nodupkey .. when it shouldn't.
e.g.
data have_w_char_id;
input id $;
cards;
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
;
run;
proc sort data=have_w_char_id (keep=id) out=c_id_nodup noduprec;
by id;
run;
proc sort data=have_w_char_id (keep=id) out=c_id_nodupkey nodupkey;
by id;
run;
data have_w_num_id;
input id;
cards;
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
;
run;
proc sort data=have_w_num_id (keep=id) out=n_id_nodup noduprec;
by id;
run;
proc sort data=have_w_num_id (keep=id) out=n_id_nodupkey nodupkey;
by id;
run;
data have_w_char_id2;
input id $ x;
cards;
1 1
2 2
3 3
4 4
5 5
6 6
7 7
1 7
2 6
3 5
4 4
5 3
6 2
7 1
1 1
2 2
3 3
4 4
5 5
6 6
7 7
1 7
2 6
3 5
4 4
5 3
6 2
7 1
;
run;
proc sort data=have_w_char_id2 (keep=id) out=c_id_nodup2 noduprec;
by id;
run;
proc sort data=have_w_char_id2 (keep=id) out=c_id_nodupkey2 nodupkey;
by id;
run;
data have_w_num_id2;
input id x;
cards;
1 1
2 2
3 3
4 4
5 5
6 6
7 7
1 7
2 6
3 5
4 4
5 3
6 2
7 1
1 1
2 2
3 3
4 4
5 5
6 6
7 7
1 7
2 6
3 5
4 4
5 3
6 2
7 1
;
run;
proc sort data=have_w_num_id2 (keep=id) out=n_id_nodup2 noduprec;
by id;
run;
proc sort data=have_w_num_id2 (keep=id) out=n_id_nodupkey2 nodupkey;
by id;
run;
Art, CEO, AnalystFinder.com
