Skip to contents

parsnip model spec for PU learning with xplus

Source: R/parsnip_registration.R
pu_xplus.Rd

Create a parsnip spec driving xplus(); use with parsnip::fit_xy()/fit(), tune::tune_grid(), or finetune::tune_sim_anneal().

Usage

pu_xplus(
  mode = "classification",
  engine = "xplus",
  backend = "glmnet",
  seed = NULL,
  max_iter = 100,
  nfolds = 10,
  trees = NULL,
  learn_rate = NULL,
  tree_depth = NULL,
  min_n = NULL,
  mtry = NULL,
  sample_size = NULL,
  loss_reduction = NULL,
  ...
)

Arguments

mode

Model mode; "classification" is the only supported value.

engine

Fitting engine; "xplus".

backend

Backend selector: a built-in name such as "glmnet" or "xgboost", an xplus_backend() object, or a constructor function. When a name is given, the boost hyperparameter arguments below map onto the matching xplus_backend_*() constructor and accept tune().

seed, max_iter, nfolds

xplus() controls forwarded verbatim.

trees, learn_rate, tree_depth, min_n, mtry, sample_size, loss_reduction

Backend constructor arguments; passed through to the matching xplus_backend_*() constructor when backend is a name.

...

Reserved.

Value

A parsnip model_spec.