Warning: Inner iterations did not coverge and an advise about model structure #36
Description
Hello- I have two loosely related question maybe you can help. A bit of background:
I have data about the forms that a parasite can take in infected mice, so each datapoint is a parasite. Parasites are collected within mice, and mice within experiments. Each mouse is infected with one of four lines of parasites. The question of interest is the effect of parasite line on the proportion of the three forms accounting for the effect of mouse and experiments. In particular, I'm interested in the difference between the WT line and each of the three KO lines. This is a summary of number of mice, lines and experiments:
Line
Experiment KO_1 KO_3 KO_7 WT
A 0 3 0 3
B 0 3 0 2
C 0 0 3 3
D 3 0 3 3
E 3 0 3 3
The model I have is:
bfitu <- mblogit(Form ~ Line, random=list(~1|Experiment/mouse_id), data=lldat)
emmeans::emmeans(bfitu, ~ Line | Form)
This gives sensible results but I get the warnings below. Should I be worried about them?
Warning: Inner iterations did not coverge - nlminb message: false convergence (8)
Iteration 1 - deviance = 12516.11 - criterion = 0.9603497Warning: Inner iterations did not coverge - nlminb message: false convergence (8)
Iteration 2 - deviance = 12430.75 - criterion = 0.05545146Warning: Inner iterations did not coverge - nlminb message: false convergence (8)
Iteration 3 - deviance = 12421.89 - criterion = 0.01194718Warning: Inner iterations did not coverge - nlminb message: false convergence (8)
Iteration 4 - deviance = 12420.05 - criterion = 0.001537364Warning: Inner iterations did not coverge - nlminb message: false convergence (8)
Iteration 5 - deviance = 12419.78 - criterion = 2.682554e-05Warning: Inner iterations did not coverge - nlminb message: false convergence (8)
Iteration 6 - deviance = 12419.75 - criterion = 6.59256e-09
converged
Second question: I'm not entirely sure that a more appropriate model would be:
bfitu2 <- mblogit(Form ~ Line, random=list(~1|Experiment, ~1|mouse_id), data=lldat)
any advise on this would be much appreciated.
The dataset is attached (data.tsv.gz)
Thanks for the great package!
sessionInfo()
R version 4.3.2 (2023-10-31)
Platform: x86_64-conda-linux-gnu (64-bit)
Running under: Ubuntu 22.04.3 LTS
Matrix products: default
BLAS/LAPACK: /home/dario/miniforge3/envs/20240220_multinomial/lib/libopenblasp-r0.3.26.so; LAPACK version 3.12.0
locale:
[1] LC_CTYPE=en_GB.UTF-8 LC_NUMERIC=C
[3] LC_TIME=en_GB.UTF-8 LC_COLLATE=en_GB.UTF-8
[5] LC_MONETARY=en_GB.UTF-8 LC_MESSAGES=en_GB.UTF-8
[7] LC_PAPER=en_GB.UTF-8 LC_NAME=C
[9] LC_ADDRESS=C LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_GB.UTF-8 LC_IDENTIFICATION=C
time zone: Europe/London
tzcode source: system (glibc)
attached base packages:
[1] stats graphics grDevices utils datasets methods base
other attached packages:
[1] ggbeeswarm_0.7.2 emmeans_1.10.0 mclogit_0.9.6 Matrix_1.6-5
[5] nnet_7.3-19 ggplot2_3.4.4 data.table_1.14.10
loaded via a namespace (and not attached):
[1] gtable_0.3.4 jsonlite_1.8.8 dplyr_1.1.4 compiler_4.3.2
[5] tidyselect_1.2.0 scales_1.3.0 yaml_2.3.8 lattice_0.22-5
[9] coda_0.19-4.1 R6_2.5.1 generics_0.1.3 MASS_7.3-60
[13] tibble_3.2.1 munsell_0.5.0 pillar_1.9.0 rlang_1.1.3
[17] utf8_1.2.4 estimability_1.4.1 cli_3.6.2 withr_3.0.0
[21] magrittr_2.0.3 grid_4.3.2 mvtnorm_1.2-4 xtable_1.8-4
[25] memisc_0.99.31.7 beeswarm_0.4.0 lifecycle_1.0.4 vipor_0.4.7
[29] vctrs_0.6.5 glue_1.7.0 farver_2.1.1 fansi_1.0.6
[33] colorspace_2.1-0 tools_4.3.2 pkgconfig_2.0.3