Load mikropml with library() in preprocess script

Sometimes preprocess_data() throws an error when calling
caret::dummyVars():

Error in get(ctr, mode = "function", envir = parent.frame()) :
  object 'contr.ltfr' of mode 'function' was not found
Calls: <Anonymous> ... model.matrix -> model.matrix.default -> contrasts -> get

This error happens sometimes when using categorical features,
but not for all datasets with categorical features.
It happened to Jay with her CRC datasets split by condition when the
plate ID column was included as a feature.

For some reason, loading either mikropml or caret with library() in
scripts fixes the problem.

Author: kelly-sovacool

code/preproc.R

@@ -1,8 +1,10 @@
 source("code/log_smk.R")
+library(mikropml)
+
 doFuture::registerDoFuture()
 future::plan(future::multicore, workers = snakemake@resources[["ncores"]])
 
 data_raw <- readr::read_csv(snakemake@input[["csv"]])
-data_processed <- mikropml::preprocess_data(data_raw, outcome_colname = "dx")
+data_processed <- preprocess_data(data_raw, outcome_colname = "dx")
 
 saveRDS(data_processed, file = snakemake@output[["rds"]])