Using array to count observations per group

[GS2]

Using SAS 9.4

My data is all character (60 variables) except for the ID and the characters are either "checked" or "unchecked" as the response.

I have an array (see below) that I am trying to count every time a variable is "checked" by ID. The data set is unique so it is only one row per ID that I am trying to count. When I run the code the count continues down each row of the data set. I would appreciate any insight into how to correct the code. Thank you

data array(drop=i);
	set check;
    array b _character_;
	by id;
    do i = 1 to dim(b);
        if b[i] =: 'C' then do;
            if first.id then count + 1;
        end;
    end;
run;

It should be 

ID      Count

7162    2

7168    4

7169    4

7170    2

etc. 

[cminard]

Maybe something like?

proc sort data=check; by id; run;
proc transpose data=check out=tcheck;
	var var1-var60;
	by id;
run;

data tcheck2;
	set tcheck;
	by id;
	if first.id then counter=1*(col1='checked');
	else counter+1*(col1='checked');
	if last.id then output;
run;

[ballardw · Accepted solution]

---

@GS2 wrote:

Using SAS 9.4

My data is all character (60 variables) except for the ID and the characters are either "checked" or "unchecked" as the response.

I have an array (see below) that I am trying to count every time a variable is "checked" by ID. The data set is unique so it is only one row per ID that I am trying to count. When I run the code the count continues down each row of the data set. I would appreciate any insight into how to correct the code. Thank you

data array(drop=i);
	set check;
    array b _character_;
	by id;
    do i = 1 to dim(b);
        if b[i] =: 'C' then do;
            if first.id then count + 1;
        end;
    end;
run;

When ever you use the construct

then count + 1;

you have told SAS implicitly to retain the value. So possibly you just need to use: then count = sum(count,1);

Use of sum this way means that you do not need to initialize a value for count.