Save results after margins

[Nicolas Orgeira]

Hi,

I'm using margins after running multi-level logistic models and would like to save the outputs to produce tables and graphs at any time, that is without needing to rerun the models and margins (as it can be quite time-consuming) days after having run the models. Do you know whether this would be feasible?

Thanks

[Jeff Pitblado (StataCorp)]

Yes, this is feasible with any estimation command that works with
margins.

See the post option in the margins documentation and
documentation for commands estimates save and estimates use.

[Nicolas Orgeira]

Thanks Jeff Pitblado (StataCorp), that's perfect! Anyway I can manipulate the estimates afterwards? I'd like to use the estimates for a coefficient plot, and report the margins as percentage change as opposed to unit change.

For example:

use https://stats.idre.ucla.edu/stat/data/hsbdemo, clear
logit honors i.female read
margins, dydx(female) post
/*
Average marginal effects                                   Number of obs = 200
Model VCE: OIM

Expression: Pr(honors), predict()
dy/dx wrt:  1.female

------------------------------------------------------------------------------
             |            Delta-method
             |      dy/dx   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
      female |
     female  |   .1514206   .0513275     2.95   0.003     .0508205    .2520208
------------------------------------------------------------------------------              
*/

estimates save "test_margins.dta", replace
estimates use "test_margins"
coefplot

I'd like to change the x-axis a report a 15% increase in probability as opposed to 0.15 if possible.

Thanks again!

[Nicolas Orgeira]

All good, I see that using the option rescale(XX) does the trick!