forde {arf}R Documentation

Forests for Density Estimation


Uses a pre-trained ARF model to estimate leaf and distribution parameters.


  oob = FALSE,
  family = "truncnorm",
  finite_bounds = FALSE,
  alpha = 0,
  epsilon = 0,
  parallel = TRUE



Pre-trained adversarial_rf. Alternatively, any object of class ranger.


Training data for estimating parameters.


Only use out-of-bag samples for parameter estimation? If TRUE, x must be the same dataset used to train arf.


Distribution to use for density estimation of continuous features. Current options include truncated normal (the default family = "truncnorm") and uniform (family = "unif"). See Details.


Impose finite bounds on all continuous variables?


Optional pseudocount for Laplace smoothing of categorical features. This avoids zero-mass points when test data fall outside the support of training data. Effectively parametrizes a flat Dirichlet prior on multinomial likelihoods.


Optional slack parameter on empirical bounds when family = "unif" or finite_bounds = TRUE. This avoids zero-density points when test data fall outside the support of training data. The gap between lower and upper bounds is expanded by a factor of 1 + epsilon.


Compute in parallel? Must register backend beforehand, e.g. via doParallel.


forde extracts leaf parameters from a pretrained forest and learns distribution parameters for data within each leaf. The former includes coverage (proportion of data falling into the leaf) and split criteria. The latter includes proportions for categorical features and mean/variance for continuous features. The result is a probabilistic circuit, stored as a data.table, which can be used for various downstream inference tasks.

Currently, forde only provides support for a limited number of distributional families: truncated normal or uniform for continuous data, and multinomial for discrete data. Future releases will accommodate a larger set of options.

Though forde was designed to take an adversarial random forest as input, the function's first argument can in principle be any object of class ranger. This allows users to test performance with alternative pipelines (e.g., with supervised forest input). There is also no requirement that x be the data used to fit arf, unless oob = TRUE. In fact, using another dataset here may protect against overfitting. This connects with Wager & Athey's (2018) notion of "honest trees".


A list with 5 elements: (1) parameters for continuous data; (2) parameters for discrete data; (3) leaf indices and coverage; (4) metadata on variables; and (5) the data input class. This list is used for estimating likelihoods with lik and generating data with forge.


Watson, D., Blesch, K., Kapar, J., & Wright, M. (2023). Adversarial random forests for density estimation and generative modeling. In Proceedings of the 26th International Conference on Artificial Intelligence and Statistics, pp. 5357-5375.

Wager, S. & Athey, S. (2018). Estimation and inference of heterogeneous treatment effects using random forests. J. Am. Stat. Assoc., 113(523): 1228-1242.

See Also

adversarial_rf, forge, lik


arf <- adversarial_rf(iris)
psi <- forde(arf, iris)

[Package arf version 0.2.0 Index]