[R] Generalized Estimating Functions
Frederico Zanqueta Poleto
fred at poleto.com
Mon May 13 13:56:19 CEST 2002
Hi Thomas,
Thank you for your help.
You are right. I've tried the OME example and the working correlation was
right.
Maybe it is a problem with Poisson, what do you think?
> desreg<-gee
(desovas~ger+pop+fec+ger*pop+ger*fec+pop*fec,id=unidade,data=desovas.ovos.in
viaveis,family=poisson,corstr="exchangeable")
[1] "Beginning Cgee S-function, @(#) geeformula.q 4.13 98/01/27"
[1] "running glm to get initial regression estimate"
[1] 0.93464328 0.02317831 0.29278443 0.21399704 0.11593571
0.05833780 -0.11513896
> summary(desreg)
GEE: GENERALIZED LINEAR MODELS FOR DEPENDENT DATA
gee S-function, version 4.13 modified 98/01/27 (1998)
Model:
Link: Logarithm
Variance to Mean Relation: Poisson
Correlation Structure: Exchangeable
Call:
gee(formula = desovas ~ ger + pop + fec + ger * pop + ger * fec +
pop * fec, id = unidade, data = desovas.ovos.inviaveis, family =
poisson,
corstr = "exchangeable")
Summary of Residuals:
Min 1Q Median 3Q Max
-4.5893464 -1.5463047 -0.4124404 1.4106536 9.2329743
Coefficients:
Estimate Naive S.E. Naive z Robust S.E. Robust z
(Intercept) 0.93464317 0.06012985 15.5437480 0.05954487 15.6964517
ger 0.02317832 0.07001297 0.3310576 0.06915388 0.3351702
pop 0.29278452 0.06565443 4.4594785 0.06505729 4.5004104
fec 0.21399711 0.07252180 2.9507967 0.07150063 2.9929402
ger:pop 0.11593572 0.07355902 1.5760911 0.07318636 1.5841166
ger:fec 0.05833782 0.07000740 0.8333093 0.07054498 0.8269592
pop:fec -0.11513904 0.07216442 -1.5955098 0.07149595 -1.6104274
Estimated Scale Parameter: 1.366923
Number of Iterations: 1
Working Correlation
[,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12]
[1,] 1 0 0 0 0 0 0 0 0 0 0 0
[2,] 0 0 0 0 0 0 0 0 0 0 0 0
[3,] 0 0 0 0 0 0 0 0 0 0 0 0
[4,] 0 0 0 0 0 0 0 0 0 0 0 0
[5,] 0 0 0 0 0 0 0 0 0 0 0 0
[6,] 0 0 0 0 0 0 0 0 0 0 0 0
[7,] 0 0 0 0 0 0 0 0 0 0 0 0
[8,] 0 0 0 0 0 0 0 0 0 0 0 0
[9,] 0 0 0 0 0 0 0 0 0 0 0 0
[10,] 0 0 0 0 0 0 0 0 0 0 0 0
[11,] 0 0 0 0 0 0 0 0 0 0 0 0
[12,] 0 0 0 0 0 0 0 0 0 0 0 0
Best regards,
--
Frederico Zanqueta Poleto
fred at poleto.com
--
"It would be possible to describe everything scientifically, but it would
make no sense; it would be without meaning, as if you described a
Beethoven symphony as a variation of wave pressure." Albert Einstein
------------- Original message follows -------------
On Sun, 12 May 2002, Frederico Zanqueta Poleto wrote:
> Hi,
>
> I'm trying to fit a marginal model via GEE but I'm getting strange
> results and few problems.
> If I set the working correlation as exchangeable I'm getting the same
> fitting when I set as independent. Comparing to SAS results it shouldn't
> happen.
> If I try to use another working correlation (like AR-M or stat_M_dep), R
> just exits without giving any error message.
We probably need an example. The `OME' example in help(gee) gives
different results using exchangeable and independence working
correlations, so it isn't a universal problem.
> Another doubt is how R estimates the scale parameter for Poisson? I get
> 1.36 in R, 1.25 in SAS (estimation by the square root of DEVIANCE/DOF)
> and 1.17 (estimation by the square root of Pearson's Chi-Square/DOF).
> I'm using R 1.4.1 but all the problems apply for S-Plus 2000 too.
This is because 1.36=1.17^2, I should think. The disperson parameter from
gee() is a variance scale factor not a standard deviation scale factor.
-thomas
Thomas Lumley Asst. Professor, Biostatistics
tlumley at u.washington.edu University of Washington, Seattle
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
More information about the R-help
mailing list