Using Label define for a variable with many observations

[Esi Hutchful]

Hello all,

I am seeking help with the following. I have a dataset that is of population projections for each county in a state. There are 58 observations for the "county" variable, currently coded according to their FIPs (Federal Information Processing Standards) county code. For example, the value 6001 corresponds to Alameda county. The actual county names are only in the data dictionary. The dataset itself only has the FIPS codes.

What I am trying to do is use the command "label define" to define a label named County so that instead of reading the value "6001" when I browse my data, I can read "Alameda." I have used "label define" before but only very simply (for example, label define Sex 0 "Female" 1 "Male"). The code I have so far is what follows:

label define County 6001 "Alameda" 6003 "Alpine" 6005 "Amador" 6007 "Butte" 6009 "Calaveras" 6011 "Colusa" 6013 "Contra Costa"

Do I have to complete this process for every single one of the 58 observations? Is there a simpler way or do I need to just import the data with the county names replacing the FIPS codes before I even begin working with Stata?

Many thanks,
Esi

[David Benson]

I'm sending this from my phone, so we'll see how this works.

I think you'll be better off just creating a table with FIPS codes and county names, and then just merging in based on the fips codes. That way you'll have both the FIPS code and the county name in your dataset should you need them in the future.

[Raahil Madhok]

As David mentioned, merge your data dictionary with your dataset. If you have a .dta file with the FIPS code and county name, just run

 merge 1:1 fips_code using datadictionary.dta

. This assumes your data is at the county-level.

Now that you have both FIPS code and county name in the same dataset, you can write

labmask fips_code, values(county_name)

to assign each FIPS code a value from its corresponding county name. You'll have install labmask as its a user-written command.

[Esi Hutchful]

Thank you! This was helpful.