Function reference

Core Models

High-level model interfaces for supervised learning and causal inference.

bart.BARTModel	Class that handles sampling, storage, and serialization of stochastic forest models for supervised learning.
bcf.BCFModel	Class that handles sampling, storage, and serialization of stochastic forest models for causal effect estimation.

Scikit-Learn Interface

stochtree models wrapped as sklearn-compatible estimators.

sklearn.StochTreeBARTRegressor	A scikit-learn-compatible estimator that implements a BART regression model.
sklearn.StochTreeBARTBinaryClassifier	A scikit-learn-compatible estimator that implements a binary probit BART classifier.

Low-Level API — Data

Data structures for custom sampling workflows.

data.Dataset	Wrapper around a C++ class that stores all of the non-outcome data used in stochtree. This includes:
data.Residual	Wrapper around a C++ class that stores residual data used in stochtree.

Low-Level API — Forest

Forest containers and inspection.

forest.Forest	In-memory python wrapper around a C++ tree ensemble object
forest.ForestContainer	Container that stores sampled (and retained) tree ensembles from BART, BCF or a custom sampler.

Low-Level API — Samplers

Sampler classes for building custom models.

sampler.ForestSampler	Wrapper around many of the core C++ sampling data structures and algorithms.
sampler.GlobalVarianceModel	Wrapper around methods / functions for sampling a “global” error variance model
sampler.LeafVarianceModel	Wrapper around methods / functions for sampling a “leaf scale” model for the variance term of a Gaussian