topic Replicate a single row of data n times in SAS Programming

Replicate a single row of data n times

DB_ECON — Mon, 19 Jul 2010 13:21:31 GMT

Suppose I have the following dataset.

id var1 var2 var3.
1 1 4 6
2 7 4 0
3 1 9 6

How do I replicate each observation (1,2, and 3) twice so that the final output looks like....

1 1 4 6
1 1 4 6
2 7 4 0
2 7 4 0
3 1 9 6
3 1 9 6

One (inefficient) way of doing this is to make two identical datasets and merge them with a set statement. Is there a more efficient way of doing this.

Thanks a lot.

Message was edited by: DB_ECON Message was edited by: DB_ECON

Re: Replicate a single row of data n times

sbb — Mon, 19 Jul 2010 13:38:23 GMT

Just hardcode two OUTPUT statements in a DATA step with your SET statement.

Scott Barry
SBBWorks, Inc.

Re: Replicate a single row of data n times

Cynthia_sas — Mon, 19 Jul 2010 14:02:12 GMT

Hi:
Regarding:
merge them with a set statement

The SET statement allows you to bring datasets together VERTICALLY, either:
1) stack datasets (dataset 1 has 3 obs, dataset 2 has 3 obs) the final dataset will have 6 obs, The first dataset listed on the SET statement will contribute the first group of observations; the dataset listed second on the SET statement will contribute the second group of observations.

2) interleave datasets -- with the above scenario, you would still end up with 6 obs, but in interleaved order.

The MERGE statement allows you to join datasets HORIZONTALLY -- so for example, if you have 3 obs in both datasets and all 3 obs in both datasets match on a BY variable, then the output dataset would have 3 observations and each observation would have the variables that came from both datasets.

But, as Scott pointed out, the simple solution is to use 2 OUTPUT statements. So you won't need to "merge with a set".

cynthia

Re: Replicate a single row of data n times

divyanksingh12 — Tue, 20 Jun 2017 07:45:06 GMT

Hi DB_ECON

You can do that very easily by just appending and then sorting them.

Suppose your dataset name is A and you need output in B

Program:-

data B;

set A A;

run;

proc sort data=B;

by ID;

run;

You will get your required result in B dataset.

Re: Replicate a single row of data n times

yifan88 — Mon, 25 Feb 2019 14:53:02 GMT

You can try the following code( as single is your original data set )

data new ;
set single;
do i=1 to 2;
output;
end;
run;
proc print data=new(drop=i) noobs;
run;