Hello,
Using a fixed effects linear model (with lm), I can get exact p-values
out of the AVOVA table, even if they are very small, eg. 1.0e-200.
Using lme (linear mixed effects) from the nlme library,
it appears that there is rounding of the p-values to zero, if
the p-value is less than about
hi,
Is there a way we can obtain
the exact p-values from lme without rounding?
use summary instead.
used commands:
library(nlme)
g-lme(value~factor(fac1)+factor(fac2)+factor(fac1):factor(fac2),data=mydataframe,random=~1|factor(fac3))
ag-anova(g)
summary(g)$tTable[,5] will provide