creating frequency table with percentage

[lillymaginta1]

data have;
infile cards ;
input ID    var1  var 2 var3 ;	 
cards;
001 1 1  0
002 0 0  1
003 0 1  0
004 1 1  0
005 0 1  0
006 1 0  0
;

I am trying to create a simple summary dataset that contains only the frequency and percentage of a few variables (all coded as 0 or 1). Output dataset: 

var_name freq percent 
var1 3 50
va2  4  66.7
var3 1  16.7

[PaigeMiller · Accepted solution]

proc summary data=have;
    var var1 var2 var3;
    output out=want(drop=_:) sum=freq1-freq3 mean=percent1-percent3;
run;

The percent will be a proportion between 0 and 1, rather than an actual percent between 0 and 100. If you want to multiply it by 100, you can do that in a DATA step or apply the PERCENT format.

[lillymaginta1]

Thank you!

[FreelanceReinh]

Hello @lillymaginta1,

With a minor variant of @PaigeMiller's solution -- using the STACKODS option of PROC MEANS and an ODS OUTPUT statement -- you can get a bit closer to your intended dataset structure:

proc means data=have stackods sum mean;
var var1--var3;
ods output summary=want(rename=(variable=var_name sum=freq mean=percent));
run;

Printed with suitable formats, dataset WANT looks almost like your sample output: 

proc format;
picture mypct (round)
0-high = "009.9" (mult=1000);
run;

proc print data=want noobs;
format freq best. percent mypct.;
run;

[tarheel13]

Proc freq already gets n and % for you.

%macro freqmac(var=);
proc freq data=have noprint;
	tables var&var/ out=freq&var;
run;

data freq&var;
	set freq&var;
	if var&var=0 then delete;
	cat="var&var";
%mend;

%freqmac(var=1)
%freqmac(var=2)
%freqmac(var=3);

data stack;
	set freq1-freq3;
	drop var1-var3;
run;

proc print data=stack;run;