Longitudinal data with unique study visits

[nanavel]

I have longitudinal data. Each participant has multiple data entries for each visit.
Unfortunately, the data does not have unique visit numbers for each participant.
For eg. Id 1 will have visit numbers 111,484,222,222, 456 
And ID 2 will have to visit no’s 890, 098, 098, 450.
I would like to write a code that can assign
e.g., Id 1 visits as 111=1 ,484=2,222=3,222=3, 456=4
And I’d 2 will have visits as 890=1, 098=2, 098=2, 450=3.

This will provide consistency for all study participants for further analysis.

Any guidance will be appreciated. Thank you.

[Reeza]

Add an enumeration/counter variable.

https://stats.idre.ucla.edu/sas/faq/how-can-i-create-an-enumeration-variable-by-groups/

---

@nanavel wrote:

I have longitudinal data. Each participant has multiple data entries for each visit.
Unfortunately, the data does not have unique visit numbers for each participant.
For eg. Id 1 will have visit numbers 111,484,222,222, 456 
And ID 2 will have to visit no’s 890, 098, 098, 450.
I would like to write a code that can assign
e.g., Id 1 visits as 111=1 ,484=2,222=3,222=3, 456=4
And I’d 2 will have visits as 890=1, 098=2, 098=2, 450=3.

This will provide consistency for all study participants for further analysis.

Any guidance will be appreciated. Thank you.

---

[nanavel]

Thank you Reeza. When I try adding an enumeration, I get it as counting variables for each datapoint per participant. For e.g. ID 1 will be numbered 1-20. Meanwhile, some of them will be same-day visits so I would want those as repeats but it does not work with this code.

proc sort data=-data have;

by id;

run;

data want;

set have;

count + 1;

by id;

if first.id then count=1;

run;

[Reeza]

Try adding the DATE to your BY statement and using FIRST.DATE instead of FIRST.ID.

There are more complicated examples in that post that illustrate how to do it within different By groupings.