Count across variables (substring)

[terjeph]

Hi,

I'd like to count the number of a specific value ('4') per individual in the second digit of a character variable (one per day) and store the count (new variable). The data set:

pid    date1   date2   date3  ...

1       142      121     141

2       111       12B    121

3       141       14A    121

Anyone? 

[Jagadishkatam]

Hope its what you are expecting

data have;
input pid    date1$   date2$   date3$;
array cnt(3) cnt1-cnt3;
array dat(3) date1-date3;
do i = 1 to 3;
if substr(dat(i),2,1)='4' then cnt(i)=1;
end;
count=sum(of cnt1-cnt3);
cards;
1       142      121     141
2       111       12B    121
3       141       14A    121
;

[terjeph · Accepted solution]

Thanks!

[RW9]

Mark the actual answer to the question as correct, not your thankyou post.  Now it seems that "Thank you" is the correct answer!

[RW9]

Post test data in the form of a datastep.

We are not here to type it in for your or guess what it should look like.

If you only have one 4 per item then simply:

data have;
  input pid date1 $ date2 $ date3 $;
  tmpstr=lengthn(compress(cats(of date:),'4','k'));
datalines;
1       142      121     141
2       111       12B    121
3       141       14A    121
;
run;