[R-sig-ME] alternative of_lmeresampler::bootstrap_ for objects of class glmmTMB and/or glmmadmb
bbolker @ending from gm@il@com
Fri Aug 31 20:56:09 CEST 2018
Can you be more specific about what you'd like to do? It looks like
the lmeresampler::bootstrap functions might be adaptable to glmmTMB
objects (i.e., not in their current form, but since they are mostly
built on accessor methods that work similarly between glmmTMB and lme4,
new generics could be rewritten without huge effort -- probably).
However ... at a quick glance, it looks like only fully parametric
bootstrapping and case-resampling will work for GLMMs rather than LMMs
(and you're probably using glmmTMB/glmmADMB because you need some sort
of exotic GLMM-type thing, not a LMM, anyway ...)
You can also do parametric boostrapping with bootMer in glmmTMB, with
the current CRAN lme4 and the parboot branch of glmmTMB, although there
are still some bugs for binomial-type models:
On 2018-08-31 06:21 AM, C. AMAL D. GLELE wrote:
> Hi, dear all.
> Is there an alternative of lmeresampler::bootstrap for objects of class
> glmmTMB and/or glmmadmb?
> In advance, thanks for your help.
> Kind regards,
> [[alternative HTML version deleted]]
> R-sig-mixed-models using r-project.org mailing list
More information about the R-sig-mixed-models