How to plot univariable coefficients below multivariable coefficients via -coefplot-

[Tiago Pereira]

Dear Statalisters,

I would like to plot univariable coefficients right below multivariable coefficients - for an easy assessment of confouding factors. I have been using -coefplot-.

This is what I have been doing:

* --------- begin example ----------------*
clear
set obs 100
gene y = rnormal(100,20)

forvalues i = 1/6 {
gene cov`i' = runiform()
}

regress y cov1-cov6
estimates store multi

forvalues i = 1/6 {
regress y cov`i'
estimates store uni`i'
}

coefplot multi uni1 uni2 uni3 uni4 uni5 uni6 , drop(_cons)
* --------- end example ----------------*

This is almost what I need. However, I would like to have two colors only (navy = multivariable, cranberry = univariable). Although I don't mind manually editing, the problem is that the spacing between estimates seem to increase from top to bottom.

Do you have any suggestions how to overcome that problem?

All the best,

Tiago

[Tiago Pereira]

eststo multivariable: regress y cov1-cov6
eststo uni1: regress y cov1
eststo uni2: regress y cov2
eststo uni3: regress y cov3
eststo uni4: regress y cov4
eststo uni5: regress y cov5
eststo uni6: regress y cov6

coefplot (uni1 \ uni2 \ uni3 \ uni4 \ uni5 \ uni6, label(univariable)) (multivariable), drop(_cons)