使用glmmod <- lme4::glmer(formula, family = Gamma(link = "identity"), data = df)计算模型后，当我要求confint(glmmod)时出现错误
> confint(glmmod)
Computing profile confidence intervals ...
Error in profile.merMod(object, which = parm, signames = oldNames, ...) :
can't (yet) profile GLMMs with non-fixed scale parameters
类似于@ Ben-Bolker在using profile and boot method within confint option, with glmer model中所述。但是，如果我按照他的建议尝试confint(glmmod,method="boot")（无论是否使用nsim），我都会收到错误消息
Error in if (const(t, min(1e-08, mean(t, na.rm = TRUE)/1e+06))) { :
missing value where TRUE/FALSE needed
In addition: Warning messages:
1: In rgamma(nsim * length(ftd), shape = shape, rate = shape/ftd) :
NAs produced
2: In bootMer(object, FUN = FUN, nsim = nsim, ...) :
some bootstrap runs failed (500/500)
然后，我尝试了两种方法：一种是按照Ben Bolker的回答建议，使用glmmTMB（也尝试了glmmADMB，但又出现了另一个错误）；另一种方法是使用glmer和confint(glmmod, method = "Wald")来确定置信区间。
我现在的问题是，glmmTMB的估计超出了glmer的Wald置信区间。此外，考虑到AIC和BIC，glmer的结果似乎比glmmTMB的结果要好。
我应该信任哪个结果？我该怎么办？
谢谢！