Generate a counter summing the occurrences of an element in various variables.

[Raphael Jud]

Hello dear Statamasters,

I met an issue trying to generate a counter for my data frame.
I want to generate the yearly number of occurrences where some country-code (gwno) appears in the various variables (here gwno_a1, gwno_a2 and gwno_b1).

Here, an example of the form of my dataset :

year gwno_a1 gwno_a2 gwno_b1
1985 750 625 2
1985 530 . 666
1985 625 . 750
1985 645 700 811
1986 630 . 750
1986 530 . 666
1986 811 . 645
1986 560 . 625

And I want to obtain something that looks like the following dataset :

year gwno count
1985 2 1
1985 530 1
1985 625 2
1985 645 1
1985 666 1
1985 700 1
1985 750 2
1985 811 1
1986 530 1
1986 560 1
1986 625 1
1986 630 1
1986 645 1
1986 666 1
1986 750 1
1986 811 1


Thank you so much for your help !

[Nick Cox]

clear 
input year gwno_a1 gwno_a2 gwno_b1
1985 750 625 2
1985 530 . 666
1985 625 . 750
1985 645 700 811
1986 630 . 750
1986 530 . 666
1986 811 . 645
1986 560 . 625
end 

bysort year : gen id = _n 
reshape long gwno, i(year id) string 
drop _j id
drop if gwno == .
contract year gwno, freq(count)
l

[Raphael Jud]

Thank you so much Nick !
That very concise piece of code will help me a lot !