[R-sig-ME] reporting model estimates values
Juan Pablo Edwards Molina
edwardsmolina at gmail.com
Tue Oct 24 15:16:26 CEST 2017
Dear glmmTMB authors & list members ,
For glmer models I use "lsmeans" or "effects" package to report the
coefficient estimated values, but any of them works with glmmTMB
objetcs:
lsmeans::lsmeans(mod, ~ factor, type = "response"); or
effects::allEffects(mod)
How would you suggest me to do that with an glmmTMB model?
Here is a short example (with only two factor levels):
library(glmmTMB)
m1 <- glmmTMB(count~ mined + (1|site),
zi=~0,
family=poisson, data=Salamanders)
summary(m1)
m2 <- glmer(count ~ mined + (1|site), family=poisson, data=Salamanders)
summary(m2)
library(lsmeans)
lsmeans(m2, ~ mined, type = "response")
library(effects)
eff.m2 <- allEffects(m2); as.data.frame(eff.m2[[1]])
lsmeans(m1, ~ mined, type = "response")
eff.m2 <- allEffects(m1); as.data.frame(eff.m1[[1]])
Thanks in advance,
Juan Edwards
[[alternative HTML version deleted]]
More information about the R-sig-mixed-models
mailing list