creating one variable with multiple levels

[Tara Boyle]

Dear all,

I currently have this data


I would like to create a variable = mobility for each ID which can take
-the value of 1 if mobility_1 = 1 if timefromop = prep
-Value of 2 if mobility_2 = 1 if timefromop = prep
-Values of 3 if mobility_3 = 1 if timefromop = prep
-A value of missing if all mobility* = 0


Therefore I can then calculate a total mobility score for just preop for each id at each timeop point
i.e for postop at 30 days, 6 months and 12 months

* Example generated by -dataex-. For more info, type help dataex

clear

input int form float timefromop str11 id byte(sex age mobility___1 mobility___2 mobility___3 selfcare___1 selfcare___2 selfcare___3)

1 1 "66607102"  0 . 0 1 0 1 0 0

1 3 ""          . . 0 0 0 0 0 0

2 1 "159309038" 0 . 0 1 0 0 1 0

3 1 "19305580"  1 . 0 1 0 1 0 0

3 3 ""          . . 0 0 0 0 0 0

4 1 "104252077" 0 . 0 1 0 1 0 0

5 1 "172656787" 1 . 0 1 0 1 0 0

6 1 "68009893"  0 . 0 1 0 1 0 0

6 3 ""          . . 0 1 0 1 0 0

7 1 "56151137"  0 . 0 1 0 1 0 0

8 1 "87923470"  0 . 0 1 0 0 1 0

9 1 "58882663"  0 . 0 1 0 1 0 0

10 1 "149272395" 1 . 0 1 0 1 0 0

11 1 "67013300"  0 . 1 0 0 1 0 0

12 1 "51911287"  0 . 0 1 0 1 0 0

end

label values timefromop r10

label def r10 1 "preop", modify

label def r10 3 "postop6M", modify

I’m not sure if I explained this clearly but I’ve created a draft dataset of what I think it should look like

* Example generated by -dataex-. For more info, type help dataex

clear

input float(timefromop id mobility)

1 66607104 2

2 66607104 .

. 66607104 .

1 15930938 2

end

label values timefromop r0

label def r0 1 "preop", modify

label def r0 2 "postop30days", modify

Appreciate if you can help me solve the problem in terms of how I should best sort out the dataset
Should I just keep preop values, work on this dataset with code and repeat for different time periods followed by append the datasets?

[Clyde Schechter]

Well, in your example data it is always the case that exactly one of the variables mobility___1 through mobility___3 equals 1. You do not say what you want to do if more than one of them equals 1, or if none of them do. As your question is incomplete, here is an incomplete answer:

gen byte mobility = .
forvalues i = 1/3 {
    replace mobility = `i' if mobility___`i' == 1 & timefromop == "preop":r10
}

This code leaves mobility as missing value if none of the mobility___* variables takes on a 1 value, and it uses the number of the last mobility___* variables to have a 1 value if there is more than one such.