angelinetsui
commented
2 years ago @melsod , I looked into the issues, I found that the UK and combined models needed to trim more. I do not know why it was not a problem in 2021 when I ran the models, but right now, both models (except the NAE model) needed to remove more random effects to reach convergence. I have ensured that this is not due to any updates in my lme4 package.
I have sequentially removed (1| labid) in the UK model and also the CDI.z_age_months random effect from (1| labid). Now the file can be knitted.
If time permits, I will look into the coding about "checking assumptions" of each model (esp for the autocorrections part). but because assumptions checking is not part of our pre-registrated methods in RR Stage 1, I will focus on my time on other parts, such as exploratory analysis and etc
Joscelinrocha
I am getting an error related to furrr:
Quitting from lines 508-528 (03a_confirmatory_analysis.Rmd) Error in lme.formula(CDI.prop ~ CDI.z_age_months + language_zone + gender + : nlminb problem, convergence error code = 1 message = iteration limit reached without convergence (10) Calls: ... withVisible -> eval -> eval -> -> lme.formula
In addition: Warning messages:
1: package 'future' was built under R version 4.1.2
2: In supportsMulticoreAndRStudio(...) :
[ONE-TIME WARNING] Forked processing ('multicore') is not supported when running R from RStudio because it is considered unstable. For more details, how to control forked processing or not, and how to silence this warning in future R sessions, see ?parallelly::supportsMulticore
3: package 'furrr' was built under R version 4.1.2
Execution halted