ppd_gamboostlss: Fit boosted 'gamlss'-type semi-parametric models

In jbrowell/ProbCast: Flexible Probabilistic Forecasting

Fit boosted gamlss-type semi-parametric models

Description

This function is a wrapper for the function gamboostLSS, which fits semi-parametric regression models for predictive distributions with up to four parameters (location, scale, shape1, shape2) via gradient boosting.

Usage

ppd_gamboostlss(
  data,
  formula,
  cv_folds = NULL,
  family = gamboostLSS::GaussianLSS(),
  cores = 1,
  pckgs = NULL,
  save_models_path = NULL,
  ...
)

Arguments

data	A data.frame containing target and explanatory variables. May optionally contain a column called "kfold" with numbered/labeled folds and "Test" for test data.
formula	A formula or named list of formulas for the location, scale, shape etc..(see ?gamboostLSS)
cv_folds	Control for cross-validation if not supplied in data.
family	A gamboosLSS family object, which is used to define the distribution and the link functions of the various parameters.
cores	the number of available cores. Defaults to one, i.e. no parallelisation, although in this case the user must still specify pckgs if applicable.
pckgs	specify additional packages required for each worker (e.g. c("data.table") if data stored as such).
save_models_path	Path to save models. Model details and file extension pasted onto this string.
...	additional arguments passed to gamboostLSS(), e.g. use control = mboost::boost_control() to specify boosting steps, shrinkage etc.

Value

A list of gamboostlss objects. Each list element corresponds to a cross-validation fold and contains a gamlss model trained on all other folds.

Author(s)

Ciaran Gilbert, [e-mail]

jbrowell/ProbCast documentation built on July 20, 2024, 1:53 p.m.