Hi Michael,
This problem should be fixed in Zelig_3.0-3 which was just released.
Please install/update this version using:
install.packages("Zelig", repos="http://gking.harvard.edu")
or
install.packages("Zelig", repos="http://gking.harvard.edu",
type="source")
if you are using mac.
Thank you for sending this bug,
Best,
Ferdi
Michael Braun wrote:
First, thanks for creating such a useful R package. I
am a new Zelig
(2.8.3) user, but I can already see its time-saving potential.
However, I am having some trouble using the bprobit model, and I hope
someone can help me. I want to model something like:
mu1 = x1 %*% b1
mu2 = x2 %*% b2
rho = f(x3 %*% b3)
where mu1 and mu2 are the means of the underlying bivariate probit, and
f() is the exp(x)-1/exp(x)+1 transformation. x1, x2 and x3 are vectors
of variables, and are all different.
The zelig command I am using is
z.out <- zelig(list(mu1 = y1~x1, mu2 = y2~x2, rho = ~x3),
model="bprobit",data=mydata)
Note that x1 is really x11+x12+x13+..., and so forth.
My problem is that when I do summary(z.out), I get coefficient estimates
for b1 and b2, but only a single intercept estimate for b3. If I change
x3 variables, I get the same estimate for b3. It is as though zelig is
ignoring all of the covariates I put on rho.
I didn't see a reference to this in the list archives, but I'm wondering
if this is a known issue. From the documentation, it looks like this
should be possible.
Thanks in advance,
Michael
--
Ferdinand Alimadhi
Manager of Statistical Programming
Harvard University
The Institute for Quantitative Social Science
(617) 496-0187
falimadhi(a)iq.harvard.edu
-
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/