Sum across observations

[Ana Oseguera]

I have a dataset with regions and an id for a specific sector. As an example I have

region id x1
1 2361 6
1 2362 10
1 2363 11
1 2364 4
2 2361 8
2 2362 23
2 2363 6
2 2364 2

I need to sum 2361+2362 for each region and keep the id code of 2361 so that
region id x1
1 2361 16
1 2363 11
1 2364 4
2 2361 31
2 2363 6
2 2364 2

I found some previous posts about how to do it, but they seem to be a bit too complicated to do for all the variables. I need to do it for all 75 variables on the dataset.

[Robert Picard]

Here's one way of doing this

* Example generated by -dataex-. To install: ssc install dataex
clear
input byte region int id byte x1
1 2361 6 
1 2362 10 
1 2363 11 
1 2364 4 
2 2361 8 
2 2362 23 
2 2363 6 
2 2364 2 
end

* verify assumptions about the data
isid region id, sort

* group both 2361 2362 codes together
replace id = 2361 if id == 2362

* reduce to one obs per region id
collapse (sum) x1, by(region id)

list, sepby(region) noobs

[Ana Oseguera]

That worked. Thank you