a list of distinct values where a second variable satisfies a certain condition

[Dan Daugaard]

I’m attempting to list distinct values where a second variable satisfies a certain condition.
I’m trying to implement the advice “After using egen, tag() to create a tagged variable, you can list tagged observations to again show the distinct combinations.” from Speaking Stata: Distinct observations, The Stata Journal (2008), 8, Number 4, pp. 557–568 (on page 11 of my pdf version of the article)

Here’s an attempt using the auto file – ie listing distinct car makes where they are foreign.

 
 * list unique make with foreign attribute
clear
sysuse auto
egen tag = tag(foreign make)
egen nvals = total(tag), by(foreign)
tabdisp foreign, cell(nvals)

but I don’t know how to set the list command to list the makes which have now been tagged?

I can see the end goal with the bysort command, eg
bysort foreign: list make
however, the eventual data file I’ll be analysing has a tens of thousands of unique variables and only a small number of distinct variables to be listed.
I hope this makes sense, thank you for reading through the problem, Dan

[Jorrit Gosens]

I'm not terribly sure what you are trying to do with the first few lineof codes (the tag==1 for all observations here), but your question

list the makes which have now been tagged

Would be:

tabdisp make if tag==1, cell(nvals)

Or if only the foreign vehicles:

tabdisp make if tag==1 & foreign==1, cell(nvals)

Is that what you were after?

[Nick Cox]

I am not clear what the problem is here beyond specifying if.

make is an identifier in the auto data in any case so

list make if foreign, noobs

lists the distinct (different) values, all of which happen to be unique (occur once only). On the terminology here, see e.g. Section 2 in https://www.stata-journal.com/sjpdf....iclenum=dm0042

The paper discusses technique. If interested in the distinct command, get the program files from .

SJ-15-3 dm0042_2  . . . . . . . . . . . . . . . . Software update for distinct
        (help distinct if installed)  . . . . . .  N. J. Cox and G. M. Longton
        Q3/15   SJ 15(3):899
        improved table format and display of large numbers of
        observations

search dm0042_2, entry sj

gives you a clickable link.

[Dan Daugaard]

Thank you your help Jorrit and Nick,
Yes Jorrit, that tabdisp command followed on naturally from my egen, tag() steps and translated easily over to the context of my data.
Thank you Nick for the references and clarification around the "unique" term - I can now see my choice of using auto wasn't a good illustration of my problem/task. I'll build a better example using a sample of my data when I encounter this issue again.
Thanks again, Dan