SAS combine two columns in a dataset - sas

Question

I got a dataset looks like this:
Category
Sub-Category
value
A
1
xx
A
2
xx
A
3
xx
A
4
xx
B
1
xx
B
2
xx
B
3
xx
B
4
xx
I want to combine the first two columns and create a new dataset with the new category(rows) and it should looks like this:
Category
Value
A
1
xx
2
xx
3
xx
4
xx
B
1
xx
2
xx
3
xx
4
xx
...
Can anyone help me with that using SAS?
Thanks in advance!

Answer 1

That seems pretty useless as a DATASET.
Why not just print the original dataset.
proc print data=have;
by category;
run;
Result:
If you did want to generate that goofy dataset you could try interleaving two copies of the original data.
data want;
set have(in=in1) have(in=in2);
by category;
if in1 then do;
if first.category then call missing(value);
else delete;
end;
if in2 then category=cats(subcategory);
drop subcategory ;
run;
Result:

Answer 2

You can do this in a data step with multiple output statements. The only tricky part is the first row: you'll need to retain the original value to create the "header" row, output, then output a second time to create the first subcategory row.
data want;
set have end=eof;
by category subcategory;
if(first.category) then do;
/* "Header" row: */
_value_ = value;
value = ' ';
output;
/* First subcategory row: */
value = _value_;
category = strip(subcategory);
output;
end;
/* Subsequent subcategory rows */
else do;
category = strip(subcategory);
output;
end;
/* Add a space between categories except the last row */
if(last.category AND NOT eof) then do;
call missing(value, category);
output;
end;
drop _value_ subcategory;
run;
Result:
category value
A
1 xx
2 xx
3 xx
4 xx
B
1 xx
2 xx
3 xx
4 xx

Answer 3

data have;
input Category $ SubCategory value $;
datalines;
A 1 xx
A 2 xx
A 3 xx
A 4 xx
B 1 xx
B 2 xx
B 3 xx
B 4 xx
;
data want;
set have;
by Category;
if first.Category then do;
v = value;
value = '';
output;
value = v;
Category = put(SubCategory, 8. -l);
output;
end;
else do;
Category = put(SubCategory, 8. -l);
output;
end;
drop SubCategory v;
run;

Answer 4

Proc REPORT can produce the desired rendering of the data set.
proc report data=have;
columns category subcategory value;
define category / order noprint;
compute before category;
line category $200.;
endcomp;
run;
Some alternative renderings
proc print data=have;
by category;
id category;
var subcategory value;
run;
proc tabulate data=have;
class category subcategory;
var value;
table
category
* subcategory
,
value
;
run;
data haveR;
set have;
by category;
if first.category then output;
output;
run;
title "Report - indent";
proc report data=haveR;
column category subcategory cat value;
define category / order noprint;
define subcategory / display noprint;
define value / display;
define cat / computed style=[just=left] 'Category/hierarchy';
compute cat / length=200;
endcomp;
compute value;
if not missing(category) then do;
priorcategory = category;
cat = category;
value = .;
end;
else do;
cat = repeat('A0'x,3) || cats(subcategory);
end;
endcomp;
run;