Reference

Estimation API Tidy functions for performing an analysis of heterogeneous treatment effects. Once a configuration has been defined (like by the Recipe API), these functions are the workhorses that perform all estimation.
attach_config()	Attach an HTE_cfg to a dataframe
make_splits()	Define splits for cross-fitting
produce_plugin_estimates()	Estimate models of nuisance functions
construct_pseudo_outcomes()	Construct Pseudo-outcomes
estimate_QoI()	Estimate Quantities of Interest
Recipe API Tidy functions for configuring a “recipe” for how to estimate heterogeneous treatment effects. This is the easiest way to get started with setting up a configuration for an HTE analysis.
basic_config()	Create a basic config for HTE estimation
add_propensity_score_model()	Add an additional model to the propensity score ensemble
add_known_propensity_score()	Uses a known propensity score
add_propensity_diagnostic()	Add an additional diagnostic to the propensity score
add_outcome_model()	Add an additional model to the outcome ensemble
add_outcome_diagnostic()	Add an additional diagnostic to the outcome model
add_effect_model()	Add an additional model to the joint effect ensemble
add_effect_diagnostic()	Add an additional diagnostic to the effect model
add_moderator()	Adds moderators to the configuration
add_vimp()	Adds variable importance information
remove_vimp()	Removes variable importance information
Model Configuration Classes to define the configuration of models to be used in the eventual HTE analysis. These are the classes which define the underlying configurations in the Recipe API. They’re most useful for advanced users who want the most granular control over their analysis, but most users will be best served by the Recipe API.
Base Class
Model_cfg	Base Class of Model Configurations
Models for displaying results These model configurations include valid standard errors and are well suited for providing usable output to be plotted or returned for formal inference.
Stratified_cfg	Configuration for a Stratification Estimator
KernelSmooth_cfg	Configuration for a Kernel Smoother
Models for nuisance functions These models are most useful for estimating nuisance functions in the course of HTE estimation. In particular, Known_cfg may be used when the propensity score is known ex ante, while SLEnsemble_cfg may be used when a SuperLearner ensemble of machine learning models should be used to estimate an unknown nuisance function.
Known_cfg	Configuration of Known Model
Constant_cfg	Configuration of a Constant Estimator
SLEnsemble_cfg	Configuration for a SuperLearner Ensemble
SLLearner_cfg	Configuration of SuperLearner Submodel
Analysis Configuration These classes configure the overall shape of the HTE analysis: essentially, how all the various components and models should fit together. They explain what models should be estimated and how those models should be combined into relevant quantities of interest. These, too, underlie the Recipe API, and should rarely need to be used directly.
Diagnostics_cfg	Configuration of Model Diagnostics
HTE_cfg	Configuration of Quantities of Interest
QoI_cfg	Configuration of Quantities of Interest
VIMP_cfg	Configuration of Variable Importance
Configuration of HTE estimands There are two configurations for HTE estimands. MCATE_cfg simply provides a one-dimensional slice of the effects over a particular dimension. It averages over all other moderators, so does not attempt to attribute heterogeneity to particular covariates. PCATE_cfg, however, does attempt to do this type of attribution. Estimation of these “partial” effects requires strong assumptions, even within the context of a randomized control trial. Essentially, one must assume that the moderator, too, is random conditional on treatment and other covariates.
MCATE_cfg	Configuration of Marginal CATEs
PCATE_cfg	Configuration of Partial CATEs
Internal Functions The remaining functions are not really useful for end-users. Documentation is provided in order to provide additional details about the internal workings of the methods.
SL.glmnet.interaction()	Elastic net regression with pairwise interactions
predict(<SL.glmnet.interaction>)	Prediction for an SL.glmnet object
calculate_ate()	Calculates a SATE and a PATE using AIPW
calculate_diagnostics()	Calculate diagnostics
calculate_rroc()	Regression ROC Curve calculation
calculate_linear_vimp()	Calculate Linear Variable Importance of HTEs
calculate_vimp()	Calculate Variable Importance of HTEs
split_data()	Partition the data into folds
Model_data	R6 class to represent data to be used in estimating a model
check_data_has_hte_cfg()	Checks that a dataframe has an attached configuration for HTEs
check_identifier()	Checks that an appropriate identifier has been provided
check_nuisance_models()	Checks that nuisance models have been estimated and exist in the supplied dataset.
check_splits()	Checks that splits have been properly created.
check_weights()	Checks that an appropriate weighting variable has been provided
fit_plugin()	Fits a plugin model using the appropriate settings
fit_plugin_A()	Fits a propensity score model using the appropriate settings
fit_plugin_Y()	Fits a T-learner using the appropriate settings
fit_effect()	Fits a treatment effect model using the appropriate settings
listwise_deletion()	Removes rows which have missing data on any of the supplied columns.