Skip to contents

Get the Default Values for the Extra Computation Arguments

Source: R/setup.R
get_extra_comp_args_default.Rd

Get the Default Values for the Extra Computation Arguments

Usage

get_extra_comp_args_default(
  internal,
  paired_shap_sampling = isFALSE(internal$parameters$asymmetric),
  semi_deterministic_sampling = FALSE,
  kernelSHAP_reweighting = "on_all_cond",
  compute_sd = isFALSE(internal$parameters$exact),
  n_boot_samps = 100,
  vS_batching_method = "future",
  max_batch_size = 10,
  min_n_batches = 10,
  max_batch_cube_size = 1e+06,
  global_loss_func = NULL
)

Arguments

internal

List. Not used directly, but passed through from explain().

paired_shap_sampling

Logical. If TRUE paired versions of all sampled coalitions are also included in the computation. That is, if there are 5 features and e.g. coalitions (1,3,5) are sampled, then also coalition (2,4) is used for computing the Shapley values. This is done to reduce the variance of the Shapley value estimates. TRUE is the default and is recommended for highest accuracy. For asymmetric, FALSE is the default and the only legal value.

semi_deterministic_sampling

Logical. If FALSE (default), then we sample from all coalitions. If TRUE, the sampling of coalitions is semi-deterministic, i.e. the sampling is done in a way that ensures that coalitions that are expected to be sampled based on the number of coalitions are deterministically included such that we sample among fewer coalitions. This is done to reduce the variance of the Shapley value estimates, and corresponds to the PySHAP* strategy in the paper Olsen & Jullum (2025).

kernelSHAP_reweighting

String. How to reweight the sampling frequency weights in the kernelSHAP solution after sampling. The aim of this is to reduce the randomness and thereby the variance of the Shapley value estimates. The options are one of 'none', 'on_N', 'on_all', 'on_all_cond' (default). 'none' means no reweighting, i.e. the sampling frequency weights are used as is. 'on_N' means the sampling frequencies are averaged over all coalitions with the same original sampling probabilities. 'on_all' means the original sampling probabilities are used for all coalitions. 'on_all_cond' means the original sampling probabilities are used for all coalitions, while adjusting for the probability that they are sampled at least once. 'on_all_cond' is preferred as it performs the best in simulation studies, see Olsen & Jullum (2025).

compute_sd

Logical. Whether to estimate the standard deviations of the Shapley value estimates. This is TRUE whenever sampling based kernelSHAP is applied (either iteratively or with a fixed number of coalitions).

n_boot_samps

Integer. The number of bootstrapped samples (i.e. samples with replacement) from the set of all coalitions used to estimate the standard deviations of the Shapley value estimates.

vS_batching_method

String. The method used to perform batch computing of vS. "future" (default), utilizes future.apply::future_apply (via the future::future package), enabling parallelized computation and progress updates via progressr::progressr. Alternatively, "forloop" can be used for straightforward sequential computation, which is mainly useful for package development and debugging purposes.

max_batch_size

Integer. The maximum number of coalitions to estimate simultaneously within each iteration. A larger number requires more memory, but may have a slight computational advantage.

min_n_batches

Integer. The minimum number of batches to split the computation into within each iteration. Larger numbers give more frequent progress updates. If parallelization is applied, this should be set no smaller than the number of parallel workers.

max_batch_cube_size

Numeric. The largest number of elements allowed in the dense per-batch array built by the gaussian, copula and empirical approaches. For gaussian and copula this array has a total of n_MC_samples * n_explain * coalitions_per_batch * n_features elements, while for empirical it is the distance array with n_train * n_explain * coalitions_per_batch elements. When a batch would exceed this, the batch size is automatically reduced (i.e. more batches are used) and a message is given. The default 1e6 keeps peak memory modest and tends to reduce runtime in high-dimensional settings, while staying far below the 32-bit indexing limit of the underlying RcppArmadillo arrays (which fails with Cube::init(): requested size is too large). Raise it to allow larger batches, or lower it to use even smaller ones.

global_loss_func

Function or NULL. Only used when scope = "global" (i.e. when computing SAGE values). The loss function used to measure the model loss when computing the SAGE values. Must take two arguments, the true response and the model prediction (in that order), and return a single numeric loss value. If NULL (default), logistic (cross-entropy) loss is used for binary responses (values in 0/1) and mean squared error loss otherwise.

Value

A list with the default values for the extra computation arguments.

References

Author

Martin Jullum