Model insights

Obtain insights from a black box model in the form of feature effects.

insights(
  mfit,
  vars,
  data,
  interactions = "user",
  hcut = 0.75,
  pred_fun = NULL,
  fx_in = NULL,
  ncores = -1
)

Arguments

mfit	Fitted model object (e.g., a "gbm" or "randomForest" object).
vars	Character vector specifying the features to get insights on.
data	Data frame containing the original training data.
interactions	String specifying how to deal with interaction effects: 'user' specify interactions in vars as "var1_var2". 'auto' automatic selection of interactions based on hcut.
hcut	Numeric in the range [0,1] specifying the cut-off value for the normalized cumulative H-statistic over all two-way interactions, ordered from most to least important, between the features in vars. Note that hcut = 0 will add the single most important interaction, while hcut = 1 will add all possible two-way interactions.
pred_fun	Optional prediction function to calculate feature effects for the model in mfit. Requires two arguments: object and newdata. See pdp::partial and this article for the details. See also the function gbm_fun in the example.
fx_in	Optional named list of data frames containing feature effects for features in vars that are already calculated beforehand, to avoid having to calculate these again. A possible use case is to supply the main effects such that only the interaction effects still need to be calculated. Precalculated interactions are ignored when interactions = "auto", but can be supplied when interactions = "user". It is important to make sure that you supply the pure interaction effects.
ncores	Integer specifying the number of cores to use. The default ncores = -1 uses all the available physical cores (not threads), as determined by parallel::detectCores(logical = 'FALSE').

Value

List of tidy data frames (i.e., "tibble" objects), containing the partial dependencies for the features (and interactions) in vars.

Examples

if (FALSE) {
data('mtpl_be')
features <- setdiff(names(mtpl_be), c('id', 'nclaims', 'expo', 'long', 'lat'))
set.seed(12345)
gbm_fit <- gbm::gbm(as.formula(paste('nclaims ~',
                               paste(features, collapse = ' + '))),
                    distribution = 'poisson',
                    data = mtpl_be,
                    n.trees = 50,
                    interaction.depth = 3,
                    shrinkage = 0.1)
gbm_fun <- function(object, newdata) mean(predict(object, newdata, n.trees = object$n.trees, type = 'response'))
gbm_fit %>% insights(vars = c('ageph', 'bm', 'coverage', 'fuel'),
                     data = mtpl_be,
                     interactions = 'auto',
                     hcut = 0.75,
                     pred_fun = gbm_fun)
gbm_fit %>% insights(vars = c('ageph', 'bm', 'coverage', 'fuel', 'bm_fuel'),
                     data = mtpl_be,
                     interactions = 'user',
                     pred_fun = gbm_fun)
}