Creating time indicators

[bgosiker]

Hello! 

I am trying to run a discrete time to event model and am trying to generate indicator variables for each respective study visit. I have people entering in at different times in a cohort (so not everyone has a visit 1 necessarily) . This would mean that the first visit for a person should have an indicator at time0 regardless of the visit number that they enter at and then each subsequent study visit will activate the next indicator variable. 

Given only the id and visit number the data I WANT generated is: 

id     visit     time0     time2     time3     time4     time5      time6      time7 

1    1     1      0      0      0      0      0      0

1    2     0      1      0      0      0      0      0   

1    3     0      0      1      0      0      0      0

1    4     0      0      0      1      0      0      0

1    5     0      0      0      0      1      0      0

1    6     0      0      0      0      0      1      0

1    7     0      0      0      0      0      0      1

2    2     1      0      0      0      0      0      0

2    3     0      1      0      0      0      0      0    

2    4     0      0      1      0      0      0      0

2    5     0      0      0      1      0      0      0

2    6     0      0      0      0      1      0      0

2    7     0      0      0      0      0      1      0

2    8     0      0      0      0      0      0      1

3    1     1      0      0      0      0      0      0

3    2     0      1      0      0      0      0      0

3    3     0      0      1      0      0      0      0

3    4     0      0      0      1      0      0      0

3    5     0      0      0      0      1      0      0

3    6     0      0      0      0      0      1      0

3    7     0      0      0      0      0      0      1

Here is the data that I have: 

Data have;
 input ID  visit ;
datalines;
1  1
1  2
1  3
1  4
1  5
1  6
1  7
2  2
2  3
2  4
2  5
2  6
2  7
3  1
3  2
3  3
3  4
3  5
3  6
;
run;

I appreciate any and all guidance 🙂 

[Jagadishkatam]

data want1;
set have;
by id visit;
time='time'||ifc(visit=1,'0',strip(put(visit,best.)));
val=1;
run;

options missing=0;
proc transpose data=want1 out=trans ;
by id visit;
id time;
var val;
run;

[bgosiker]

Thanks for this step! 

When I run this, it generates a time variable that is just the time combined with the visit number. I altered this a bit and generated a count variable so that the first visit for each person would register as time0 and so on (code below).  I think my issue was mainly how to get the data into the format I had shown initially where it is a bunch of indicator variables for the number visit they are at. 

data want; set have;
	count+1; 
	by id visit; 
	if first.id then count=1;
run; 

[KachiM · Accepted solution]

The data set requires to be presorted by ID.

The time0 does not match with time2. Therefore time1 to time7 is used instead. Here is Data Step.

data want;
   do i = 1 by 1 until(last.id);
      set have;
      by id;
      array t[*] time1 - time7 (7*0);
      t[i] = 1;
      output;
      t[i] = 0;
   end;
drop i;
run;

[Ksharp]

Data have;
 input ID  visit ;
datalines;
1  1
1  2
1  3
1  4
1  5
1  6
1  7
2  2
2  3
2  4
2  5
2  6
2  7
3  1
3  2
3  3
3  4
3  5
3  6
;
run;
data temp;
set have;
val=1;
run;

proc transpose data=temp out=temp1(drop=_name_) prefix=time ;
by id visit;
id visit;
var val;
run;

proc stdize data=temp1 out=want missing=0 reponly;
var _numeric_;
run;