[Andrew Musau]

See the options -omitted- and -baselevels- in

help coefplot

[Noemi Seng]

Dear Andrew Musau ,

thank you very much. That solved my problem. And also apologies, I should have looked into the helpfile first on my own.

Best wishes
Noemi

[Noemi Seng]

Dear Andrew Musau

I have another, coefplot-related question: I want to do a plot that shows the coefficients for the same variable but estimated with 5 different models. Each model is restored by

estimates use model1.ster

estimates store model1

(for the model 1 to 5).

I already managed to do

coefplot sector1 sector2 sector3 sector4 sector5, ///
>          keep(lndistw) ///
>          xline(0) ///
>          ciopts(recast(rcap) color(blue)) ///
>          msymbol(O) ///
>          ytitle("Coefficient of lndistw") ///
>          xtitle("Model") ///
>          xlabel(1 "sector1" 2 "sector2" 3 "sector3" 4 "sector4" 5 "sector5", angle(0) labsize(medium)) ///
>          vertical

However, I want the coefficients and confidence intervals nicely spread over the whole range of the x-axis and I want each of them labelled with the name of the estimation model on the x-axis.

Is this possible with coefplot?

Best
Noemi

[Andrew Musau]

I would advise you to start a new thread and include a reproducible example, even if it includes two or three models. While this thread concerns coefplot, #18 loosely fits the topic.

[Noemi Seng]

Dear Andrew Musau

thanks for the advice. I opened a new thread here: https://www.statalist.org/forums/for...imation-models

I hope it is okay that I used the data set you used before in this thread.