Merging two continuous variables in the same dataset

[Shivani Gaiha]

Hi,

I have two variables: 1) age of initiating cigarette use with values from 10-23 and 2) age of initiating hookah use with values from 11-19. Also, these variables are not mutually exclusive and so people could have used both products at different ages. For example, a person could have first-tried a cigarette at 10 years and hookah at 13. I want to combine these variables so that I can create a new variable on age of initiating tobacco substances with values that reflect both variables. I therefore, do not want the variables to simply add up the values. I want the variable to show the lowest age of initiation for each age.


ex.

ageofciguse ageofhookahuse
16 15
15 16
10 17
12 22
11 11
14 19
15 16
19 18
16 14

Ideally, I would like to create a new variable

ageoftobaccouse
15
15
10
12
11
14
15
18
14

Next, how can I do this for more than two variables?

[daniel klein]

generate newvar = min(oldvar1, oldvar2, ...)

Perhaps you need to think about missing values.

Best
Daniel

[Shivani Gaiha]

Is there a way to skip missing values from the min, and count only numeric values?

[daniel klein]

I do not understand what you mean by "skip missing values"; it seems that min() does that by default

. display min(42,., 73)
42

Also, see egen's rowmin() function.

Best
Daniel