[R-sig-ME] Problems with convergence

[Ken Beath]

The following code shows that there are convergence problem messages where
there is a problem with convergence. The profiling shows that the maximum
found is not the correct one. This is simulated data for a binary
meta-analysis with fixed effect for study and random effect for treatment.

library(lme4)

thedata <- structure(list(nEvents=c(10L,53L,17L,18L,22L,6L,16L,
14L,13L,18L,15L,19L,52L,19L,8L,16L,50L,8L,9L,4L,
26L,45L,18L,20L,5L,16L,18L,7L,3L,19L,30L,26L,66L,
23L,29L,18L,72L,25L,9L,2L),total=c(200,200,200,200,
200,200,200,200,200,200,200,200,200,200,200,200,200,
200,200,200,200,200,200,200,200,200,200,200,200,200,
200,200,200,200,200,200,200,200,200,200),trt=c(0,
0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,
1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1),id=structure(c(1L,
2L,3L,4L,5L,6L,7L,8L,9L,10L,11L,12L,13L,14L,15L,
16L,17L,18L,19L,20L,1L,2L,3L,4L,5L,6L,7L,8L,9L,
10L,11L,12L,13L,14L,15L,16L,17L,18L,19L,20L),.Label=c("1",
"2","3","4","5","6","7","8","9","10","11","12","13",
"14","15","16","17","18","19","20"),class="factor")),.Names=c("nEvents",
"total","trt","id"),row.names=c(NA,40L),class="data.frame")

glmer1<-glmer(cbind(nEvents,total-nEvents)~trt+factor(id)+(0+trt|id),data=thedata,family=binomial)

# while glmer has problems with component 9 it is 8 with a problem profile
# I've use devtol so the discrepancy is printed
prof.glmer1<-profile(glmer1,which=8,devtol=1.0e-3)


-- 

*Ken Beath*
Lecturer
Statistics Department
MACQUARIE UNIVERSITY NSW 2109, Australia

Phone: +61 (0)2 9850 8516

Building E4A, room 526
http://stat.mq.edu.au/our_staff/staff_-_alphabetical/staff/beath,_ken/

CRICOS Provider No 00002J
This message is intended for the addressee named and may...{{dropped:9}}