topic Re: sas data in SAS Programming

sas data

ganeshsas764 — Tue, 22 Mar 2016 17:55:21 GMT

hi friends ,

any body clearify my doute ,

data set " A" have 100 observations(inputdataset), "B" output dataset,

in dataset " A" ,every 5th observation should come to data set "B" ,

let me clearify me?

Davanden — Tue, 22 Mar 2016 18:07:23 GMT

One way to do this would be:

DATA B (DROP=count);
SET A;
RETAIN count 1;
IF count = 5 THEN DO;
OUTPUT;
count = 1;
END;
ELSE
count =count + 1;
RUN;

mbuchecker — Tue, 22 Mar 2016 18:15:38 GMT

or more simply:

data b;

set a;

if mod(_n_,5) = 0 then output;

run;

PeterHobart — Wed, 23 Mar 2016 09:05:17 GMT

The previous examples read every row then output every 5th.

Fine for a one-off run or a small sample.

This is more efficient technique which only reads every 5th row but it's slightly more code.

data b;

do i = 1 to rows by 5;

set a point=i nobs=rows;

output;

end;

stop;

run;

PeterHobart — Wed, 23 Mar 2016 09:14:00 GMT

One more option, if you have SAS/Stat.

this will generate a randon 20% sample

proc surveyselect data=a out=b
rate=0.2;
run;
quit;

FreelanceReinh — Wed, 23 Mar 2016 23:16:18 GMT

And if you have SAS/STAT 13.1 or higher, you can use PROC SURVEYSELECT also to literally select every 5th observation (without any random component):

proc surveyselect data=a out=b
method=sys(start=5) rate=0.2;
run;