Deal seriously with the random effects for freq models

[DominiqueMakowski]

No description provided.

[strengejacke]

What do you mean? Is it related to #57?

[DominiqueMakowski]

Pretty much yeah doing my stats I ran into some problems and it made me feel like it's worth regiving it a look, cause from what I remember previously it was possible to get predictions from lmer models with new data without random effects and now it's not

[strengejacke]

You can set the variable of the random effects to NA. This is what I did for glmmTMB. Then you get predictions on the population level.

[strengejacke]

modelbased/R/visualisation_matrix.R

Line 62 in ea26c3e

visualisation_matrix.glmmTMB <- function(x, target = "all", length = 10, factors = "reference", numerics = "mean", preserve_range = FALSE, standardize = FALSE, standardize_robust = FALSE, reference = x, na.rm = TRUE, random = TRUE, include_response = FALSE, ...) {