when lme() model contain a categorical variable, coefs() won't run properly

[iisnot]

when lme() model contain a categorical variable, coefs() won't run properly

here is the sample code for repetition in version 2.3.0.1

library(dplyr)
library(nlme)

iris = iris |> 
  mutate(Category2 = as.character(round(Sepal.Length,0)))
mod1 = lme(Sepal.Width ~ Petal.Length + Petal.Width + Category2, 
           random = ~1|Species, data = iris)
mod2 = psem(mod1)
summary(mod2)

it will return if (random) {conditions >1

i think this is caused by the function model.frame(), it pass data to random

[iisnot]

i also have a question. why a model contains a character or factor variable, the standardized path coefficients of other continues variables will not return.

[leagenty]

hello, same issue here, did you find a way to adress it ?

[kendall-jefferys]

I am also running into some issues with categorical variables.

I have mixed effect model with lme().

When I run the models including the fixed effect of sampling method (factor with 2 categories: "area based sampling" and "time based sampling"), I get this error. "Error in if (random) { : the condition has length > 1"

When I take the categorical variable out, the piecewiseSEM works.

I read the chapter on categorical variables in the online book but am still unsure how to address this.