Help generating region variable

[Jason Browen]

Dear all,

I am using Stata 16, on mac. I want to create a region variable based on my dummy variables west, ne, midwest, and south, but I am not sure how to do so. south is my reference category. I included the commands I used to construct my dummy variables along with a snapshot of my dataset.

generate west = 0

replace west =1 if 2.state | 3.state | 5.state | 6.state | 12.state | 13.state | 27.state | 29.state | 32.state | 38.state | 45.state | 48.state | 51.state

generate ne = 0

replace ne =1 if 7.state | 20.state | 30.state | 22.state | 31.state | 33.state | 39.state | 40.state | 46.state

generate midwest = 0

replace midwest =1 if 14.state | 15.state | 16.state | 17.state | 23.state | 24.state | 26.state | 28.state | 35.state | 36.state | 42.state | 50.state

generate south = 0

replace south =1 if 1.state | 4.state | 8.state | 10.state | 11.state | 18.state | 19.state |21.state | 25.state | 37.state | 34.state | 41.state | 43.state | 44.state | 47.state | 49.state | 48.state | 9.state

Thank you in advance for your help



Jason Browen

[Nick Cox]

Already answered in your earlier thread,

Please don’t ask the same question in different threads.

Please do read the FAQ Advice to see advice on screenshots.

[Jason Browen]

Sorry about that I did not know

[William Lisowski]

Instead of generating four dummy variables, generate your region variable directly by modifying your existing code, and you won't need to generate dummy variables because you can use factor variable notation in your models. Note the use of the inlist() function to simplify the if clause.

generate region = .
// west
replace region = 1 if inlist(state,2,3,5,6,12,13,27,29,32,38,45,48,51)
// northeast
replace region = 2 if inlist(state,7,20,30,22,31,33,39,40,46)
// midwest
replace region = 3 if inlist(state,14,15,16,17,23,24,26,28,35,36,42,50)
// south 
replace region = 4 if inlist(state,1,4,8,10,11,18,19,21,25,37,34,41,43,44,47,49,48,9)
label define REGION 1 "West" 2 "Northeast" 3 "Midwest" 4 "South"
label values region REGION
// make sure nothing was missed
assert region<.

Untested code. Please do read the advice on giving usable sample data with the dataex command rather than pictures of data.