version 2.1.0

The ‘build_aggtree’ function now requires a different set of arguments that allow better control on training-honest sample split.

version 2.0.3

Updated vignettes. Improved LATEX output.

version 2.0.2

Fixed a bug that occurred when using only one covariate.

version 2.0.1

Now the inference_aggtree function can estimate asymmetric bias-corrected and accelerated 95% confidence intervals for the GATEs based on their bootstrap distributions.

Also, improved automatic check of leaf composition. Functions that estimate the GATEs now:

• Raise an error if one or more leaves contain only one observation;
• Raise an error if differences in mean outcomes is the chosen estimator and one or more leaves contain only treated or only control units.

version 2.0.0

Updated hypothesis testing procedure. Improved LATEX output.

version 1.0.0

First public release