topic Re: Dividing the observations in to columns in SAS Programming

Dividing the observations in to columns

allurai0412 — Wed, 01 May 2013 18:02:24 GMT

hi,

i have some question......i have 100 observations.............. i want to assign every 10 observations to new column

shall i try with firstobs , obs or

mod (_n_,10).......??

i have data like this .

NAV:

.78 (100 the observation)

want the cloulmns as below

1_10 2_10 3 _10 4_10 5_10..........................

Regards

Allu

Re: Dividing the observations in to columns

DBailey — Wed, 01 May 2013 18:05:08 GMT

What are you wanting the value to be in the new column? The sum of something, count of something? or is it that you would ultimately want your data like this:

1 2 3 4 5 6 7 8 9 10

11 12 13 14 15 16 17 18 19 20

21 22 23 24 25 26 27 28 29 30

Re: Dividing the observations in to columns

allurai0412 — Wed, 01 May 2013 18:15:21 GMT

yes ...i need the Observation values in cloumn wise ...............

1_10 2_10 3_10............

1 11

2 12

3 13

... .

. .

10 20

Regards

Allu

Re: Dividing the observations in to columns

DBailey — Wed, 01 May 2013 19:00:45 GMT

how about;;

data want;

set have;

if _N_<10 then ColName='1_10';

else if _N_ < 20 then ColName='2_10';

else ColName='10_10';

run;

then proc transpose on colname

Re: Dividing the observations in to columns

Haikuo — Wed, 01 May 2013 20:08:55 GMT

If this is what you want, if your RAM is large enough to hold your table in whole.

data have;

do nav=1 to 99;

output;

end;

run;

data _null_;

call symputx('obs', ceil(nobs/10));

stop;

set have nobs=nobs;

run;

data want;

do i=1 to &obs ;

do j=1 to 10;

set have end=last;

array t(10,&obs.) _temporary_;

array _(&obs.);

t(j,i)=nav;

if last then leave;

end;

do j=1 to 10;

do i=1 to &obs;

_(i)=t(j,i);

end; output;

end;

stop;

keep _:;

run;

Haikuo

Re: Dividing the observations in to columns

Ksharp — Thu, 02 May 2013 02:24:57 GMT

Self-merge skill is suited for your question, If your dataset is large ,using a macro variable to wrap it.

data want;

merge have(firstobs=1 obs=10 rename=(nav=nav1))

have(firstobs=11 obs=20 rename=(nav=nav2))

..........

;

run;

Ksharp