[R] gee models summary

[Renaud Lancelot]

This inconvenience does not occur with function geese in package 
geepack. If you're using function gee in package gee, extract the 
components you want from the fitted object:

 > library(gee)
 > library(MASS)
 > data(OME)
 > fm <- gee(cbind(Correct, Trials-Correct) ~ Loud + Age + OME, id = ID,
+            data = OME, family = binomial, corstr = "exchangeable")
[1] "Beginning Cgee S-function, @(#) geeformula.q 4.13 98/01/27"
[1] "running glm to get initial regression estimate"
[1] -5.90984676  0.15516993  0.01876291 -0.07569691 -0.31660542

 > summ.fm <- summary(fm)
 > str(summ.fm)
List of 11
  $ call               : language gee(formula = cbind(Correct, Trials - 
Correct) ~ Loud + Age +      OME, id = ID, data = OME, family = 
binomial, corstr = "exchangeable")
  $ version            : chr "gee S-function, version 4.13 modified 
98/01/27 (1998)"
  $ nobs               : int 1097
  $ residual.summary   : Named num [1:5] -0.959  0.551  2.041  3.121 12.024
   ..- attr(*, "names")= chr [1:5] "Min" "1Q" "Median" "3Q" ...
  $ model              :List of 3
   ..$ link  : chr "Logit"
   ..$ varfun: chr "Binomial"
   ..$ corstr: chr "Exchangeable"
  $ title              : chr "GEE:  GENERALIZED LINEAR MODELS FOR 
DEPENDENT DATA"
  $ coefficients       : num [1:5, 1:5] -5.9010  0.1551  0.0185 -0.0418 
-0.2856 ...
   ..- attr(*, "dimnames")=List of 2
   .. ..$ : chr [1:5] "(Intercept)" "Loud" "Age" "OMEhigh" ...
   .. ..$ : chr [1:5] "Estimate" "Naive S.E." "Naive z" "Robust S.E." ...
  $ working.correlation: num [1:30, 1:30]  1.0000 -0.0234 -0.0234 
-0.0234 -0.0234 ...
  $ scale              : num 1.25
  $ error              : chr "Error code was 0"
  $ iterations         : int 4
  - attr(*, "class")= chr "summary.gee"

 > summ.fm$coefficients
                Estimate  Naive S.E.     Naive z Robust S.E.    Robust z
(Intercept) -5.90100336 0.336455218 -17.5387482 0.231463368 -25.4943295
Loud         0.15507565 0.007513314  20.6401142 0.005321880  29.1392592
Age          0.01851319 0.003561640   5.1979385 0.003305525   5.6006793
OMEhigh     -0.04183516 0.160288285  -0.2609995 0.152182352  -0.2749015
OMElow      -0.28563252 0.131496124  -2.1721745 0.117510247  -2.4307031

or better:

 > coef(summ.fm)
                Estimate  Naive S.E.     Naive z Robust S.E.    Robust z
(Intercept) -5.90100336 0.336455218 -17.5387482 0.231463368 -25.4943295
Loud         0.15507565 0.007513314  20.6401142 0.005321880  29.1392592
Age          0.01851319 0.003561640   5.1979385 0.003305525   5.6006793
OMEhigh     -0.04183516 0.160288285  -0.2609995 0.152182352  -0.2749015
OMElow      -0.28563252 0.131496124  -2.1721745 0.117510247  -2.4307031

Best,

Renaud

Dean Sonneborn a écrit :
> I'm running some GEE models but when I request the summary(pcb.gee) all 
> I get are rows and rows of intercorelations and they fill up the screen 
> buffer so I can not even scroll back to see what else might be in the 
> summary. How do I get the summary function to NOT print the 
> intercorrelations?
> Thanks,
> 

-- 
Renaud Lancelot
Directeur Adjoint chargé des Affaires Scientifiques
Deputy Director for Scientific Affairs

Département EMVT du CIRAD, TA 30/B
Campus International de Baillarguet
34398 Montpellier Cedex 5 - France
Tel.  +33 (0)4 67 59 37 17
Secr. +33 (0)4 67 59 39 04
Fax   +33 (0)4 67 59 37 95