Generate grouping variable based on various nominal variables

[Guest]

Dear community,

I'm working with household data containing various nominal or ordinal variables such as household type, income group, or location (see data example below)
I now want to put these households together in different groups, resulting in one group per possible combination of variable levels.
Is there a way to do this automatically without having to use various loops which would be tedious?

Thanks a lot,

Guest

* Example generated by -dataex-. To install: ssc install dataex
clear
input double(hhtyp2 hheink_gr1 RegioStaR7 bus28)
 3 11 77  3
 2 14 73  3
 2 10 77  3
 1  5 77 95
 3  4 73  3
 3  5 71  2
 3  9 72 95
 4  7 71  3
 4  5 77  3
 4  7 74  1
 4  6 72  1
 3  9 72  3
 2  4 72  1
 3  5 76  3
 2  7 72  1
 2  9 73  3
 3  7 75  1
 4  5 74  3
 3  5 76  1
95  5 76 95
 3  3 73  1
 2  7 76 95
 3  3 76 95
 2  9 74  5
 2  5 71 95
 3  5 77  2
 3  9 73  1
 2  7 77 95
 3  6 74  3
 3  7 72  2
 2  7 73 95
 3  6 72  1
 3  6 77  2
 4  4 71  1
 3  8 73  1
 2  9 76 95
 3  9 73  1
 3  7 71 95
 3 11 73  1
 4  5 73  2
 3  6 73  4
 3  7 73  1
 4  9 76  3
 3 13 71 95
 3  5 77  1
 3  3 76 95
 3  4 77  6
 3  7 76  1
 3 11 77 95
 4  4 72  2
 4  7 72 95
 4  4 72  1
 3 15 76  3
 2 11 74 95
 4  5 73  2
 3  5 77  4
 4  4 73  2
 4  5 77 95
 4  4 77  4
 2  5 76  1
 2  9 73  1
 4  5 73  1
 2  8 73 95
 4  8 73  1
 4  3 71  2
 4  4 77  2
 3 15 72  2
 3  3 77 95
 4  5 73 95
 2 11 73  1
 3 11 73 95
 3  9 77  5
 3  6 73 95
end

[Øyvind Snilsberg]

egen group = group(hhtyp2 hheink_gr1 RegioStaR7)

[Felix Bittmann]

Something like:

egen test = group(var1 var2 var3)
tabstat var1 var2 var3, by(test)

[Guest]

Nice, thanks a lot to you two!

[Nick Cox]

See also https://www.stata-journal.com/articl...article=dm0034 for more detailed discussion.

[Guest]

Thanks a lot!