Creating variable name from concatenation of a character string and a numeric array value

[Malthe]

Hi there and in advance thanks a lot for your help! I'm using SAS Enterprise Guide 7.15.

I have a dataset with patients, baseline-dates and scoring-dates. For each patient i have the same baseline-date and multiple scoring-dates (>20). I want to see whether the scoring-dates are within certain periods of time from the baseline-date: 1 year, 2 years, 3 years ... - 100 years;

Instead of writing multiple lines of codes, i am trying to boil it down to a more manageable size.

Here is a data-example with five scoring dates: 

data one;
length patient_id $ 8;
format baseline_date score_date date9.;
input patient_id baseline_date score_date;
cards;
Patient1 0 200
Patient1 0 400
Patient1 0 600
Patient1 0 800
Patient1 0 1000
;

run;

Here is my code:

data scoring_within_year_x;
set one;
array a_withinyear [5] (1 2 3 5 10 100); 

/* For simplicity boiled down to smaller fixed array, these are the values i need the most */ 

do i = 1 to dim(a_withinyear);
if intnx('year',baseline_date,a_withinyear[i],'same' >= score_date then 

/* Here i have my first problem, i want to name the new variable as a concatenation of a string and the value of the array - something like this */

cats('Score_within_',a_withinyear[i],'_year') = 1

/* Here i have my second problem. I want to output multiple variables to the same observation as shown underneath */ 
end;
run; 

So ideally my output for the 4th line of the original data would look like this:  

Patient_id | Baseline_date | Score_date | Score_within_1_year | Score_within_2_year | Score_within_5_year | etc...

Patient1              0                     800                         1                                      1                                0 

I hope you can help me 🙂

[yabwon · Accepted solution]

Hi,

try Proc Transpose:

data one;
length patient_id $ 8;
format baseline_date score_date date9.;
input patient_id baseline_date score_date;
cards;
Patient1 0 200
Patient1 0 400
Patient1 0 600
Patient1 0 800
Patient1 0 1000
;
run;

data tmp;
set one;
array a_withinyear [6] _temporary_ (1 2 3 5 10 100) ; 

/* For simplicity boiled down to smaller fixed array, these are the values i need the most */ 

do i = 1 to dim(a_withinyear);
  if intnx('year',baseline_date,a_withinyear[i],'same') >= score_date then 
    do;
      varname = cats('Score_within_',a_withinyear[i],'_year');
      flag = 1;
      output;
    end;

end;
run; 
proc sort data=tmp;
  by patient_id baseline_date score_date varname;
run; 

proc transpose data = tmp out = scoring_within_year_x(drop = _name_);
  by patient_id baseline_date score_date;
  var flag;
  id varname;
run;

Bart

[Malthe]

Great solution, and quick as well! Thanks! 

 I actually solved it myself before seeing your answer using a 'call execute - datastep-macro' instead, but i think i like your solution better. It must be quicker than to read and write to the same data-set many times 🙂