Multiple response variable analysis According to Total and Sex

[Akter]

Hi All, I'm struggling to generate the below output table "Output" from the below made dataset 'Have'. Your help will be highly appreciated. Cat, Dog and Birds are multiple response variable for each Subject ID. Want to generate output table as below  output table (Response Category by Total and Sex). 

Data Have;

Id     Cat   Dog   Bird   SEX
01     1       0       1         M
02     0       1       1         F
03     1       1       0         M
04     0       1       1         F
05     1       0       0         M 

Output table: 

Response	Total			Male			Female
	N	n	%	N	n	%	N	n	%
Cat	5	4	80%	3	3	100%	2	0	0
Dog	5	3	60%	3	1	33.30%	2	2	100
Bird	5	3	60%	3	1	33.30%	2	2	100

N=Total Obs (0 and 1), n=# of 1 (one)

Really appreciate your help!

[ballardw · Accepted solution]

---

@Akter wrote:

Hi All, I'm struggling to generate the below output table "Output" from the below made dataset 'Have'. Your help will be highly appreciated. Cat, Dog and Birds are multiple response variable for each Subject ID. Want to generate output table as below  output table (Response Category by Total and Sex). 

Data Have;

Id     Cat   Dog   Bird   SEX
01     1       0       1         M
02     0       1       1         F
03     1       1       0         M
04     0       1       1         F
05     1       0       0         M 

Output table: 

Response	Total			Male			Female
	N	n	%	N	n	%	N	n	%
Cat	5	4	80%	3	3	100%	2	0	0
Dog	5	3	60%	3	1	33.30%	2	2	100
Bird	5	3	60%	3	1	33.30%	2	2	100

N=Total Obs (0 and 1), n=# of 1 (one)

Really appreciate your help!

---

This should get you started:

Data Have;
input Id $     Cat   Dog   Bird   SEX $ ;
datalines;
01     1       0       1         M
02     0       1       1         F
03     1       1       0         M
04     0       1       1         F
05     1       0       0         M 
;

proc tabulate data=have;
   class sex;
   var cat dog bird;
   table cat dog bird,
         (all='Total' Sex=' ')*(n='N' sum='n'*f=best5. mean='%'*f=percent7.)
         /box='Response'
         ;
run;

Default behavior for order of class variables means the F comes before M. A custom format could be used to create text of Female and Male from values of F and M.

The ='xxx' after a statistic or variable is an override of the default label. Sum of 1/0 coded numeric is the number of 1 and mean is the percent of 1 values. The *f= is used to assign a format to the statistic value. Default would typically be 2 decimals, which wouldn't make the mean look much like a percentage.

[Akter]

Thank you so much for your help! I really appreciate your help!