SAS Array procedure

[Ngkl2001]

[dcruik]

Below is a way to do it using a macro and the Transpose Procedure and then merging the transposed data sets together.  It should get you what you are looking for in the output data.  If arrays are required, that can be adjusted too.

data have;
input ID$ Time X Y Z;
datalines;
1 10 1 2 3
1 11 4 5 6
1 12 7 8 9
2 13 10 20 30
2 14 . . .
2 15 40 . 50
3 16 15 . .
3 17 25 36 37
3 18 35 36 37
;
run;

proc sort data=have;
by ID;
run;

%macro transpose(var,n);
%if &n^=3 %then %do;
proc transpose data=have prefix=&var out=want_&var (drop=_NAME_);
by ID;
var &var;
run;
%end;
%else %do;
proc transpose data=have prefix=&var out=want_&var (drop=_NAME_);
by ID;
var &var;
run;

data want;
merge want_X want_Y want_&var;
run;
%end;
%mend;
%transpose(X,1)
%transpose(Y,2)
%transpose(Z,3);

[Ngkl2001]

thank you for your help and good answer. however, marco, proc transpose are not allowed. i can only use array data step to do it. sorry .

[Ngkl2001]

here is my raw data file, in my google drive
https://drive.google.com/open?id=0BzqllIWlzJosWGZSdmkzZ0RIeDA

[FreelanceReinh]

I think you can find the solution in this article (see Example 1 there).

[Ngkl2001]

thank you very much. let me study it first. really appreciate your help

[Ngkl2001]

libname rawfile 'c:\';
run;
proc sort data=rawfile.man;
by id;
run;

data ROWOBS;
set rawfile.man;
by id;
array x_var(3) X1-X3 ;
array y_var(3) Y1-Y3 ;
array z_var(3) Z1-Z3;
if first.id then do i = 1 to 3;
x_var(i) = .;
y_var(i) = .;
z_var(i) = .;
end;
j = time;
x_var(j) = x;
y_var(j) = y;
z_var(j) = z;
if last.id then output;
retain x1-x3 y1-y3 z1-z3;
drop i j time x y z;
run;

[Reeza]

Here's a good walk through for future reference.

http://www.ats.ucla.edu/stat/sas/modules/widetolong_data.htm