R: unexpected behaviour of function (probably due to environmental scope of formula object within function)

Question

Although a long time user of R, I have little experience to write complex programs in R. During one of my first tries, I experience a unexpected behavior of my code. Namely, if I try to run a function which I have defined my self, a wrapper around a gamlss regression model adaptation, the function fails. running the code within the function line by line interactively works just fine. Weirdly enough, once I have run the code interactively once, the function shows the expected behavior.

Here a minimal reproducible example:

First I load the required packages

    library(data.table)
    library(gamlss)

I define my wrapper function

    fun.wrapper <- function(dt, cores){
      all_vars <- colnames(dt)[colnames(dt) > "b"]
      scope_vars <- all_vars[1:ceiling(length(all_vars)/2)]
      formula_full <- as.formula(paste0("a ~ ", paste(all_vars, collapse = " + ")))
      mod0 <- gamlss(formula = formula_full, data = dt, family = "NO", trace = FALSE)
      droper <- drop1All(object = mod0, scope = scope_vars, print = FALSE, parallel = ifelse(cores>1, "multicore", "no"), ncpus = cores)
      return(droper)
    }

I initialize a random data set

    data <- data.table(
      x = rbinom(size = 1, prob = 0.3, n = 10),
      y = rbinom(size = 1, prob = 0.1, n = 10),
      z = rbinom(size = 1, prob = 0.2, n = 10),
      a = rnorm(n = 10)^2)

and I apply the function to the data, which fails with an error

    res_f1 <- fun.wrapper(dt = data, cores = 1)

if I run the code from the wrapper line by line, the results is as expected

    dt <- data
    cores <- 1
    all_vars <- colnames(dt)[colnames(dt) > "b"]
    scope_vars <- all_vars[1:ceiling(length(all_vars)/2)]
    formula_full <- as.formula(paste0("a ~ ", paste(all_vars, collapse = " + ")))
    mod0 <- gamlss(formula_full, data = dt, family = "NO", trace = FALSE)
    res_i1 <- drop1All(object = mod0, scope = scope_vars, print = FALSE, parallel = ifelse(cores>1, "multicore", "no"), ncpus = cores)

Once I have done this, I can run also successfully the wrapper function

   res_f2 <- fun.wrapper(dt = data, cores = 1)

I tried to several, mostly trial and error driven attempts to resolve this. I read about lexical scoping and closures since I strongly assume that the above drop1All call access the wrong data. I tried to assign and <<- the respective variables to the global environment. none of which did the trick.

Therefore I would be very happy if one of you could point me to a solution for this problem. thanks in advance.

sessionInfo()
R version 4.2.1 (2022-06-23)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 20.04.4 LTS

Matrix products: default
BLAS:   /usr/lib/x86_64-linux-gnu/blas/libblas.so.3.9.0
LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.9.0

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=de_AT.UTF-8        LC_COLLATE=en_US.UTF-8    
 [5] LC_MONETARY=de_AT.UTF-8    LC_MESSAGES=en_US.UTF-8   
 [7] LC_PAPER=de_AT.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=de_AT.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] parallel  splines   stats     graphics  grDevices utils     datasets 
[8] methods   base     

other attached packages:
[1] gamlss_5.4-12     nlme_3.1-162      gamlss.dist_6.0-5 MASS_7.3-58      
[5] gamlss.data_6.0-2 data.table_1.14.8

loaded via a namespace (and not attached):
[1] compiler_4.2.1  Matrix_1.5-3    survival_3.5-5  grid_4.2.1     
[5] lattice_0.20-45