create recipe from data

Usage

prepare_ml_recipe(
  data,
  prep_recipe = NULL,
  corr_method = "pearson",
  corr_use = "pairwise.complete.obs",
  thres_list = NULL,
  step_list = NULL,
  vars_imp_ignore = c(".trt"),
  vars_fct_expl_na = NULL,
  vars_ordinalscore = NULL,
  vars_keep_corr = NULL,
  level_other,
  one_hot,
  log_base
)

Arguments

data: raw data set to create recipe for
prep_recipe: if NULL, recipe will be created
corr_method, corr_use: defaulting to corr_method "pearson" and corr_use "pairwise.complete.obs"
thres_list, step_list: named list objects collecting all threshold values and step selection info, resp. please refer to the documentation of the thres_* and prep_step_* arguments in prepare_ml() for detailed documentation and list entry names.
vars_imp_ignore, vars_fct_expl_na, vars_ordinalscore, vars_keep_corr: vars relevant to steps. see prepare_ml() for details.
level_other: name for the "other"-category in recipes::step_other(). defaults to "other".
one_hot, log_base: see prepare_ml()

Value

a named list with entries containing

the prepared recipe
info on steps included in the recipe
a list of relevant variables
a list of thresholds used
high_corr a tibble listing correlations above thres_corr. NULL if step_list$prep_step_corr = FALSE.

create recipe from data

Usage

Arguments

Value

See also