How to generate a categorical variable for all combinations of a list?

[Greg Saldutte]

Hello,

I previously received help generating an indicator variable that would take the value of all possible combinations of a list. In order that the name of the indicator would correspond with what it represented, I used this command:

gen byte f  = sex==2
gen byte n  = bpl<=99
gen byte w  = race==1 & !inlist(hispan,1,2,3,4)
gen byte y  = age<=25
gen byte s  = inlist(marst,3,4,5,6)
  
forval f=0/1 {
forval n=0/1 {
forval y=0/1 {
forval w=0/1 {
forval s=0/1 {
  gen cat_f`f'_n`n'_y`y'_w`w'_s`s'= ///
  (f==`f')*(n==`n')*(y==`y')*(w==`w')*(s==`s')
}
}
}
}
}

Where f = female, n = native, w = white, y = age<=25, and s = single. Now, instead of 31 indicators, I'm trying to generate one categorical variable that will take the corresponding value for all observations. My intutition is to do the following:

generate ct = 1

and then use the egen command, but having trouble finding out how I would correctly do this.

Thank you.

[Marcos Almeida]

Maybe - egen - with - group() - would be helpful.

[Clyde Schechter]

I think this is unnecessarily complicated. See -help egen- and look for the -group()- function.

egen ct = group(f n y w s)

Added: Crossed with #2 which gives the same advice.

[Greg Saldutte]

Thank you Marcos and Clyde. You were both right.