UCLA Academic Technology Services HomeServicesClassesContactJobs
Help the Stat Consulting Group by giving a gift             
Loading

Stata FAQ
How can I graph the results of the margins command? (Stata 11)

use http://www.ats.ucla.edu/stat/data/hsbdemo, clear
anova write prog##female math read

/* plot adjusted cell means */
margins prog#female, asbalanced
mat b=r(b)'
mat prg=(1\2\3)#(1\1)
mat fem=(1\1\1)#(0\1)
mat c=prg,fem,b
svmat c, names(c)

/* plot prog by female */
twoway (connect c3 c1 if c2==0)(connect c3 c1 if c2==1), ///
       xlabel(1 2 3) legend(order(1 "female=0" 2 "female=1")) ///
       xtitle(prog) ytitle(ajdusted cell means) ///
       name(prog_by_female, replace)

/* plot female by prog */      
twoway (connect c3 c2 if c1==1)(connect c3 c2 if c1==2)(connect c3 c2 if c1==3), ///
       xlabel(0 1) legend(order(1 "prog=1" 2 "prog=2" 3 "prog=3")) ///
       xtitle(female) ytitle(ajdusted cell means) ///
       name(female_by_prog, replace)

use http://www.ats.ucla.edu/stat/data/logitcatcon, clear
logit y i.f##c.s, nolog

/* plot probabilities by categorical variable */
margins f, at(s=(20(2)70)) vsquish
mat b=r(b)'
mat at=r(at)
mat at=at[1...,3]#(1\1)
mat p=at,b
svmat p, names(p)
generate p3 = ~mod(_n,2) in 1/52
twoway (line p2 p1 if p3==0)(line p2 p1 if p3==1), ///
       legend(order(1 "f=0" 2 "f=1")) xtitle(variable s) ///
       ytitle(probability) name(probability, replace)


/* difference in probabilities */
margins, dydx(f) at(s=(20(2)70)) vsquish
mat b=r(b)'
mat at=r(at)
mat at=at[1...,3]
local nrow=rowsof(at)
mat b=b[`nrow'+1...,1]
mat se=r(V)
mat se=se[`nrow'+1...,`nrow'+1...]
mat se=vecdiag(cholesky(diag(vecdiag(se))))'
mat d=at,b,se
mat lis d
svmat d, names(d)
gen ul = d2 + 1.96*d3
gen ll = d2 - 1.96*d3
twoway (rarea ul ll d1, color(gs13) lcolor(gs13)) ///
       (line d2 d1), yline(0) legend(off) ///
       xtitle(variable s) ytitle(difference in probability) ///
       name(difference, replace)

How to cite this page

Report an error on this page or leave a comment

UCLA Researchers are invited to our Statistical Consulting Services
We recommend others to our list of Other Resources for Statistical Computing Help
These pages are Copyrighted (c) by UCLA Academic Technology Services


The content of this web site should not be construed as an endorsement of any particular web site, book, or software product by the University of California