Suppressing some variables in coefplot when using multiple regression models

[Guest]

Hi!

I want to create a coefplot which compares unadjusted and adjusted cox regression models for several outcome variables. However, I want to exclude some of the covariates included in the adjusted model from the coefficient plot - is this possible? In other words, i want the coefficient plot to display the change in HR from the unadjusted --> adjusted models only for the independent variable in my study. In my adjusted models I have adjusted for (amongst many other things) gender which is not necessary to include in the coefplot - could this be removed? I have tried using "keep", however it does not generate the desired outcome. I get the following message:

(simple_model: no coefficients found, all dropped, or none kept)
(adjusted_model: no coefficients found, all dropped, or none kept)

So the output does not change by including the keep command into the code

my code is:
coefplot simple_model adjusted_model, keep(indepvar) bylabel(" Coefplot")||, xline(1) eform byopts(compact cols(2)) baselevels


grateful for help! thank you

[Andrew Musau]

-keep()- or -drop()- should work, you need to show a reproducible example where it fails. Note that coefplot is from SSC (FAQ Advice #12).