Replies: 1 comment 2 replies
-
If you previously installed a development version of the repository - try reinstalling. There was a but in the repeated fitting procedure that messed things up a bit. If that is not the cause of the instability you are witnessing, it would be helpful to get some more details. I am not quite sure what you expect me to comment on for your code: at least half of the code is for GLMMs fitted with lme4/glmmTMB, and thus not applicable to gllvm. Or am I missing something here? |
Beta Was this translation helpful? Give feedback.
2 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi everyone,
Related to #189, I have been having difficulties getting stable output from models including LVs, but models that include random row effects seem more easily reproduced. While I'm still not entire sure how to go about obtaining output similar to
getResidCov()
+getResidCor()
that only work with LVs, I was also thinking it'd be helpful to see if I could at least produce an R2 for random row effects + fixed effect models.Digging into the supplementary code of Nakagawa and Schielzeth 2013 and Nakagawa et al. 2017, I think I have adapted a solution for
gllvm
, but I'm not sure if there are other considerations for the multivariate context or whether I'm accessing the appropriategllvm
parameters/outputs.Does this seem right? It seems like it can generate pretty high R2 for the spider data. Also, it seems like the log-normal approximation creates slightly higher R2.
Beta Was this translation helpful? Give feedback.
All reactions