Merge pull request #135 from StochasticTree/model-config-refactor

Refactor model parameters into "config" objects to future-proof low-level interface

Author: andrewherren

.github/workflows/r-test.yml

@@ -39,7 +39,7 @@ jobs:
 
       - name: Create a CRAN-ready version of the R package
         run: |
-          Rscript cran-bootstrap.R 0 0
+          Rscript cran-bootstrap.R 0 0 1
       
       - uses: r-lib/actions/check-r-package@v2
         with:

DESCRIPTION

@@ -1,5 +1,5 @@
 Package: stochtree
-Title: Stochastic tree ensembles (XBART and BART) for supervised learning and causal inference
+Title: Stochastic tree Ensembles (XBART and BART) for Supervised Learning and Causal Inference
 Version: 0.1.0
 Authors@R: 
   c(
@@ -10,7 +10,11 @@ Authors@R:
     person("Jingyu", "He", role = "aut"),
     person("stochtree contributors", role = c("cph"))
   )
-Description: Stochastic tree ensembles (XBART and BART) for supervised learning and causal inference.
+Description: Flexible stochastic tree ensemble software. Robust implementations of 
+    Bayesian Additive Regression Trees (Chipman, George, McCulloch (2010) <doi:10.1214/09-AOAS285>) 
+    for supervised learning and (Bayesian Causal Forests (BCF) Hahn, Murray, Carvalho (2020) <doi:10.1214/19-BA1195>) 
+    for causal inference. Enables model serialization and parallel sampling 
+    and provides a low-level interface for custom stochastic forest samplers.
 License: MIT + file LICENSE
 Encoding: UTF-8
 Roxygen: list(markdown = TRUE)

LICENSE

@@ -1,2 +1,2 @@
-YEAR: 2024
-COPYRIGHT HOLDER: stochtree authors
+YEAR: 2025
+COPYRIGHT HOLDER: stochtree contributors

LICENSE.md

@@ -1,6 +1,6 @@
 # MIT License
 
-Copyright (c) 2024 stochtree authors
+Copyright (c) 2023-2025 stochtree authors
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

NAMESPACE

@@ -28,7 +28,9 @@ export(createCppRNG)
 export(createForest)
 export(createForestDataset)
 export(createForestModel)
+export(createForestModelConfig)
 export(createForestSamples)
+export(createGlobalModelConfig)
 export(createOutcome)
 export(createPreprocessorFromJson)
 export(createPreprocessorFromJsonString)

NEWS.md

@@ -1,3 +1,16 @@
 # stochtree 0.1.0
 
-* Initial CRAN submission.
+* Initial release on CRAN.
+* Support for sampling stochastic tree ensembles using two algorithms: MCMC and Grow-From-Root (GFR)
+* High-level model types supported:
+    * Supervised learning with constant leaves or user-specified leaf regression models
+    * Causal effect estimation with binary, continuous, or multivariate treatments
+* Additional high-level modeling features:
+    * Forest-based variance function estimation (heteroskedasticity)
+    * Additive (univariate or multivariate) group random effects
+    * Multi-chain sampling and support for parallelism
+    * "Warm-start" initialization of MCMC forest samplers via the Grow-From-Root (GFR) algorithm
+    * Automated preprocessing / handling of categorical variables
+* Low-level interface:
+    * Ability to combine a forest sampler with other (additive) model terms, without using C++
+    * Combine and sample an arbitrary number of forests or random effects terms

R/bart.R

@@ -556,12 +556,12 @@ compute_leaf_indices_cpp <- function(forest_container, covariates, forest_nums)
   .Call(`_stochtree_compute_leaf_indices_cpp`, forest_container, covariates, forest_nums)
 }
 
-sample_gfr_one_iteration_cpp <- function(data, residual, forest_samples, active_forest, tracker, split_prior, rng, feature_types, cutpoint_grid_size, leaf_model_scale_input, variable_weights, a_forest, b_forest, global_variance, leaf_model_int, keep_forest, pre_initialized) {
-  invisible(.Call(`_stochtree_sample_gfr_one_iteration_cpp`, data, residual, forest_samples, active_forest, tracker, split_prior, rng, feature_types, cutpoint_grid_size, leaf_model_scale_input, variable_weights, a_forest, b_forest, global_variance, leaf_model_int, keep_forest, pre_initialized))
+sample_gfr_one_iteration_cpp <- function(data, residual, forest_samples, active_forest, tracker, split_prior, rng, feature_types, cutpoint_grid_size, leaf_model_scale_input, variable_weights, a_forest, b_forest, global_variance, leaf_model_int, keep_forest) {
+  invisible(.Call(`_stochtree_sample_gfr_one_iteration_cpp`, data, residual, forest_samples, active_forest, tracker, split_prior, rng, feature_types, cutpoint_grid_size, leaf_model_scale_input, variable_weights, a_forest, b_forest, global_variance, leaf_model_int, keep_forest))
 }
 
-sample_mcmc_one_iteration_cpp <- function(data, residual, forest_samples, active_forest, tracker, split_prior, rng, feature_types, cutpoint_grid_size, leaf_model_scale_input, variable_weights, a_forest, b_forest, global_variance, leaf_model_int, keep_forest, pre_initialized) {
-  invisible(.Call(`_stochtree_sample_mcmc_one_iteration_cpp`, data, residual, forest_samples, active_forest, tracker, split_prior, rng, feature_types, cutpoint_grid_size, leaf_model_scale_input, variable_weights, a_forest, b_forest, global_variance, leaf_model_int, keep_forest, pre_initialized))
+sample_mcmc_one_iteration_cpp <- function(data, residual, forest_samples, active_forest, tracker, split_prior, rng, feature_types, cutpoint_grid_size, leaf_model_scale_input, variable_weights, a_forest, b_forest, global_variance, leaf_model_int, keep_forest) {
+  invisible(.Call(`_stochtree_sample_mcmc_one_iteration_cpp`, data, residual, forest_samples, active_forest, tracker, split_prior, rng, feature_types, cutpoint_grid_size, leaf_model_scale_input, variable_weights, a_forest, b_forest, global_variance, leaf_model_int, keep_forest))
 }
 
 sample_sigma2_one_iteration_cpp <- function(residual, dataset, rng, a, b) {