Using Mgen Command and Issues with Variable Label

[Roman Johnson]

I am trying to do a post-estimation margins command, and everything looks well with my output, but I am not sure why Stata is telling me that Males doesn't exist when I'm just trying to label the variable. I am not sure if using the dataex command is appropriate or not here so forgive me for not including in this general post.


. mgen, at(numcigs_r=(.5(5)60) sex=1) atmean stub(Males) ///
> predlabel(Logit prediction for males)

Predictions from: margins, at(numcigs_r=(.5(5)60) sex=1) atmean predict(pr)

Variable Obs Unique Mean Min Max Label
--------------------------------------------------------------------------------
Malespr1 12 12 .2586882 .1080438 .4576168 Logit prediction for ...
Malesll1 12 12 .0849188 .0598954 .1058421 95% lower limit
Malesul1 12 12 .4324576 .1561922 .8524183 95% upper limit
Malesnumci~r 12 12 28 .5 55.5 numcigs_r
--------------------------------------------------------------------------------

Specified values of covariates

1. 2. 3. 1. 2.
sex race race race hrstv_r religion religion
----------------------------------------------------------------------------
1 .0809524 .097619 .0190476 13.45952 .8285714 .0309524

1. 2. 3. 1. 1. 2. 3.
friends friends friends unfrien~y exercise exercise exercise
-----------------------------------------------------------------------------
.1761905 .247619 .5380952 .3833333 .3047619 .2571429 .25

. label var Males "prediction for male"
variable Males not found
r(111);

[Carole J. Wilson]

You've created several variables that begin with Males: Malespr1, Malesll1, Malesul1... You need to be specific about which of those variables you want to label. I assume:

label var Malespr1 "prediction for male"

[Roman Johnson]

Thank you. That worked.