Dear zelig-users,
I have a few questions regarding the use of plot.ci. And if somebody is proficient with
the boolplot function (boolean) it would be great.
I am trying to produce two comparable graphs, showing the variation of the predicted
propability of Y over the full range of one indipendent variable, including the confidence
interval. Once using a boolean regression and once using a plain probit.
Since I have just started to use zelig, I am in a bit of trouble, trying to make that
work.
Now the first question:
Is it possible to change the illustration of the confidence interval? I would need a
connected line for the actual values and two lines indicating the confindence interval.
answer.zelig <- zelig(altruis2 ~ prosocial + ask1 + awelloff2 + rooms + neig +
resis, data = rd, model = "probit")
summary(answer.zelig)
prorange <- seq(0,1,by=0.01)
x.1 <- setx(answer.zelig, prosocial=prorange)
s.out <- sim(answer.zelig, x = x.1)
plot.ci(s.out)
Or: Is it easier to change the shape of the CI in a boolplot?
My third question also deals with the boolplot function. If I set the implemented
histogram to "false", the graph does not connect the values anymore. Does anyone
have an idea if and how to change that?
answer.zelig <- zelig(bp, data = rd, model = "boolean", link =
"probit", method = "BFGS")
boolplot(answer.zelig, CI=95, variable = "prosocial", truehist=F,
plot.both = F, delta = 0, yscale=c(0, 1))
Thank you all in advance for any ideas.
Best regards,
Malte Reichelt
(University of Mannheim)
--
NEU: FreePhone - kostenlos mobil telefonieren und surfen!
Jetzt informieren:
http://www.gmx.net/de/go/freephone
-
Zelig Mailing List, served by Harvard-MIT Data Center
Send messages: zelig(a)lists.gking.harvard.edu
[un]subscribe Options:
http://lists.gking.harvard.edu/?info=zelig
Zelig program information:
http://gking.harvard.edu/zelig/