Skip to contents

LightGBM Learner

Source: R/learners.R
lnr_lightgbm.Rd

A wrapper for lightgbm::lgb.train() for use in nadir::super_learner().

Usage

lnr_lightgbm(
  data,
  formula,
  weights = NULL,
  nrounds = 100,
  objective = NULL,
  verbose = -1,
  ...
)

Arguments

data

A dataframe to train a learner / learners on.

formula

A regression formula to use inside this learner.

weights

Optional observation weights.

nrounds

Number of boosting iterations.

objective

Optional objective. For binary Super Learner fits, nadir supplies objective = "binary" through outcome_type_dependent_args. If unspecified, defaults to "regression".

verbose

(default: -1) Verbosity level passed to lightgbm; -1 suppresses lightgbm's messages, larger values (0, 1, 2) show progressively more output.

...

Additional named parameters passed into the params list of lightgbm::lgb.train(), e.g. learning_rate, num_leaves, min_data_in_leaf.

Value

A prediction function that accepts newdata and returns a numeric vector of predictions (probabilities of the outcome being 1 when objective = "binary"), one for each row of newdata.

Details

{lightgbm} does not have a formula interface, so internally this learner constructs a numeric design matrix via stats::model.matrix.lm() (as in lnr_xgboost) and trains on that.

Note that min_data_in_leaf defaults to 1 here (rather than the {lightgbm} default of 20) to account for the potential of very small splits in cross-fitting; users can override this by passing min_data_in_leaf explicitly.

See also

learners

Examples

lnr_lightgbm(mtcars, mpg ~ hp + wt, nrounds = 10)(mtcars)
#>  [1] 20.68292 20.81318 21.43194 20.94345 18.04507 18.79409 16.31907 22.89741
#>  [9] 21.85530 19.05462 19.05462 17.68684 17.58914 17.58914 13.77891 15.17925
#> [17] 15.17925 25.61125 27.94516 27.94516 21.43194 17.10065 18.04507 15.66774
#> [25] 19.51054 25.61125 25.61125 26.80534 17.29605 19.83620 16.77499 20.81318

lnr_lightgbm(mtcars, am ~ cyl + disp + hp,
  objective = "binary",
  nrounds = 10
)(mtcars)
#>  [1] 0.7890306 0.7890306 0.7890306 0.1466812 0.1466812 0.1466812 0.3588241
#>  [8] 0.1466812 0.1466812 0.1466812 0.1466812 0.1466812 0.1466812 0.1466812
#> [15] 0.1466812 0.1466812 0.1466812 0.7890306 0.7890306 0.7890306 0.1466812
#> [22] 0.1466812 0.1466812 0.3588241 0.1466812 0.7890306 0.7890306 0.7890306
#> [29] 0.3588241 0.7890306 0.7890306 0.7890306