[R] Converting SAS Code

[Andrew Harmon]

I have no problem setting up my mixed model, or performing anova or lsmeans
on my model’s outputs. However, performing lsd mean separation is giving me
fits.



So I do not have a problem when using two-way anova model. When using the
code:

fit.yield.add <- lm(data = ryzup, Yield ~ Rep + Nitrogen + Treatment)

LSD.test(fit.yield.add, trt = "Nitrogen", alpha = 0.1, console = TRUE)



It works beautifully. However when I run a mixed model:



yield.lmer <- lmer(data = ryzup, Yield ~ Nitrogen + Treatment +
(1|Rep/Nitrogen), REML = FALSE)

LSD.test(yield.lmer, trt = "Nitrogen", alpha = 0.1, console = TRUE)



It gives me fits. It produces errors like:



Error in as.data.frame.default(x[[i]], optional = TRUE) :

  cannot coerce class "structure("merModLmerTest", package = "lmerTest")"
to a data.frame


Do you have any suggestions for that?

Thanks

	[[alternative HTML version deleted]]