How do I loop through parts of the observations and impute missing values

[Ian Wang]

Dear Statalist,

My data looks like this

---

id	hypertension	visit
1	.	1
1	0	2
2	.	1
2	.	2
2	1	3
3	.	1
4	.	1
5	0	1
5	1	2
5	.	3

(hypertension 1=yes;0=no; . =missing value)(Every ID had 1-3 visits)

---

I'm trying to impute all missing values according to the following rule.
"If an ID has hypertension, this ID must have hypertension in the future."
ex: ID1 didn't have hypertension on his second visit, so he didn't have hypertension on his first visit either.
ex: ID5 had hypertension on his second visit, so he must had hypertension on his third visit as well.

There are hundreds of ID in my data, so I'm trying to use loop to solve the problem.
I found something like increment might be related.
Can anyone give me more information?
Thanks!

[Rich Goldstein]

this would have been easier if you had used -dataex-; please see the FAQ; the following code does what I think you want (though your rules do not cover all possibilities):

bys id (visit): replace hy=hy[_n-1] if hy[_n-1]==1 & hy==. & id==id[_n-1]
gsort id -visit
by id: replace hy=hy[_n-1] if hy[_n-1]==0 & hy==. & id==id[_n-1]
sort id visit

this gives the following result:

. li, clean noo

    id   hy   visit  
     1    0       1  
     1    0       2  
     2    .       1  
     2    .       2  
     2    1       3  
     4    .       1  
     5    0       1  
     5    1       2  
     5    1       3

as you can see, no loop is needed

[Ian Wang]

I tried and it worked!
Thank you.

[Nick Cox]

See e.g. https://www.stata.com/support/faqs/d...issing-values/