summary.lmList issues

[bbolker]

summary.lmList can fail when some groups have singular fits. This is really an upstream problem in nlme, since we copy the summary methods for our own use ...

p2 <- subset(Pixel, Dog %in% c(7,8,9))
p3 <- subset(Pixel, Dog %in% c(7,8))  ## without problematic case
library("nlme")
L1 <- lmList(pixel~day+I(day^2)|Dog, p2)
try(summary(L1))
L2 <- lme4::lmList(pixel~day+I(day^2)|Dog, p2)
try(summary(L2))