Generic Machine Learning Inference
Performs BLP regression
Performs CLAN
Performs GATES regression
Generic Machine Learning Inference
Combine several GenericML objects
Single iteration of the GenericML algorithm
Accessor function for the best learner estimates
Accessor function for the BLP generic target estimates
Accessor function for the CLAN generic target estimates
Accessor function for the GATES generic target estimates
Evaluate treatment effect heterogeneity along CLAN variables
Estimate the two lambda parameters
Calculate lower and upper median
Plot method for a "GenericML" object
Print method for a "BLP_info" object
Print method for a "CLAN_info" object
Print method for a "GATES_info" object
Print method for a GenericML object
Print method for a "heterogeneity_CLAN" object
Propensity score estimation
Baseline Conditional Average
Conditional Average Treatment Effect
Partition a vector into quantile groups
Setup function for diff arguments
Set up information for a GenericML() plot
Setup function for stratified sampling
Setup function for vcov_control arguments
Setup function controlling the matrix in the BLP or GATES regres...
Check if user's OS is a Unix system
Generic Machine Learning Inference on heterogeneous treatment effects in randomized experiments as proposed in Chernozhukov, Demirer, Duflo and Fernández-Val (2020) <arXiv:1712.04802>. This package's workhorse is the 'mlr3' framework of Lang et al. (2019) <doi:10.21105/joss.01903>, which enables the specification of a wide variety of machine learners. The main functionality, GenericML(), runs Algorithm 1 in Chernozhukov, Demirer, Duflo and Fernández-Val (2020) <arXiv:1712.04802> for a suite of user-specified machine learners. All steps in the algorithm are customizable via setup functions. Methods for printing and plotting are available for objects returned by GenericML(). Parallel computing is supported.