Fit individual parboost component using mboost

Description

Internal function to fit mboost model on a subset of the data

Usage

1
2
3
parboost_fit(subsample_indices, data = NULL, path_to_data,
  data_import_function, preprocessing, seed, formula, baselearner, family,
  control, tree_controls, cv, cores_cv = detectCores(), folds, stepsize_mstop)

Arguments

subsample_indices

A numeric vector containing the indices of the subsample

data

A data frame containing the variables in the model. It is recommended to use path_to_data instead for IO efficiency. Defaults to NULL

path_to_data

A string with the path to the data.

data_import_function

What function should be used to import the data?

preprocessing

Optional preprocessing function to apply to the data passed from parboost

seed

Set a seed for reproducible results.

formula

Formula for mboost.

baselearner

Character string determining the type of base learner.

family

mboost family

control

mboost control

tree_controls

party control

cv

Cross-validate?

cores_cv

Number of cores to use during cv.

folds

Number of folds to use for cv.

stepsize_mstop

Stepsize used for optimizing mstop.

Details

Fits a mboost model on each subset of the data

Value

The fitted submodel and its predictions

Author(s)

Ronert Obst