RE: st: RE: marginsplot is acting differently

[Richard Williams <richardwilliams.ndu@gmail.com>]

At 05:58 AM 12/31/2011, Philip Ryan wrote:
> <<>>
>
> Richard,
>
> Further to our discussion on the behaviour of -marginsplot-, try this
> (apologies in advance if line endings are zapped):
>
> webuse nhanes2f, clear
>
> logit diabetes i.black i.female age, nolog
>
> margins, dydx(black female) at(age=(20 30 40 50 60 70)) vsquish
> marginsplot, noci
>
> margins r.black  r.female, at(age= (20 30 40 50 60 70)) vsquish
> marginsplot, noci
>
>
> The first -margins/marginsplot- sequence is as you tried, and yields the
> unwanted base level lines.  The second is the alternative discussed on pages
> 1113-1114 of the manual and _should_ give "an identical graph" (para 1 page
> 1114), but this syntax yields a graph that omits the unwanted lines.
>
> I don't know whether the first syntax's production of the base level lines
> is a feature or a bug, but it doesn't appear to be consistent with (my
> reading of) the manual.
>
> Phil

Good catch Phil! Here is the code from that section of the manual, 
and, despite what the manual says, the two graphs are clearly NOT 
identical. The 2nd one has the same weird problem that I had, i.e. it 
presents a nonsensical line for the base category that is the same as 
the x-axis.

use http://www.stata-press.com/data/r12/nhanes2, clear
logistic highbp sex##agegrp##c.bmi
margins r.sex, at(bmi=(10(1)65))
marginsplot, xlabel(10(10)60) recast(line) recastci(rarea)
margins, at(bmi=(10(1)65)) dydx(sex)
marginsplot, xlabel(10(10)60) recast(line) recastci(rarea)

Given the apparent contradiction with the manual, it is even harder 
for me to believe that this is a "feature" rather than a bug. 
Further, I know I was getting the graph I wanted and expected back in 
September, and I can't think of any good reason that you would want 
to include base-level lines when they are the exact same as the 
X-axis. I also prefer the way the margins output is formatted with my 
preferred syntax, although, as the manual notes, such preferences are 
discipline-specific.

Nonetheless, unless or until my preferred syntax starts working the 
way I want it to, it is nice to know there is such a relatively 
painless work-around. Thanks!


-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
HOME:   (574)289-5227
EMAIL:  Richard.A.Williams.5@ND.Edu
WWW:    http://www.nd.edu/~rwilliam

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/