Hello,
I want to join table by id
Table A
id date price
1 20170101 10
1 20170102 12
2 20180401 8
2 20180402 5
Table B
id date price
1 20170103 15
2 20180405 4
want A and B
id date price
1 20170101 10
1 20170102 12
1 20170103 15
2 20180401 8
2 20180402 5
2 20180405 4
There are at least a half dozen ways to do this.
If you data sets are actually sorted by the Id and date and there are not any of the same dates in A and B then
data want;
merge A
B
;
by id date;
run;
is about the simplest.
If you have duplicates of Date for a given Id then you may want:
data want;
set A
B
;
run;
proc sort data=want;
by id date;
run;
There are at least a half dozen ways to do this.
If you data sets are actually sorted by the Id and date and there are not any of the same dates in A and B then
data want;
merge A
B
;
by id date;
run;
is about the simplest.
If you have duplicates of Date for a given Id then you may want:
data want;
set A
B
;
run;
proc sort data=want;
by id date;
run;
This isn't a join, as far as I understand SQL notation and syntax. This can be achieved in SQL using a UNION but it requires another pass to get the order correct. But SAS can interleave the records; I was taught the term collate, I digress. The datasets A and B I presume are properly sorted.
data want;
set a b;
by id date;
run;
April 27 – 30 | Gaylord Texan | Grapevine, Texas
Walk in ready to learn. Walk out ready to deliver. This is the data and AI conference you can't afford to miss.
Register now and save with the early bird rate—just $795!
Learn the difference between classical and Bayesian statistical approaches and see a few PROC examples to perform Bayesian analysis in this video.
Find more tutorials on the SAS Users YouTube channel.
Ready to level-up your skills? Choose your own adventure.