NEWS.md

In ddml: Double/Debiased Machine Learning

ddml 0.3.0

• Implements one-way clustered inference.
• Increases defaults for sample_folds and cv_folds to 10.
• Fixes typo in auxiliary_X arguments.

ddml 0.2.2

• Changes ddml::ols() default to const=TRUE.
• Adds probability forest compatibility to ddml::mdl_ranger().
• Adds propensity score trimming option to ddml::ddml_ate(), ddml::ddml_att(), and ddml::ddml_late().
• Fixes ATE and LATE scores.
• Fixes output of ddml::print.summary.ddml_plm and ddml::print.summary.ddml_ate (#57).

ddml 0.2.1

• Fixes permuted residuals returned by ddml::crossval (#54).

ddml 0.2.0

• Adds support for the average treatment effect on the treated estimator.
• Adds support for local average treatment effect estimation with perfect compliance or perfect non-compliance.
• Adds support for custom ensemble weights.
• Adds article on integration with the did package.
• Adds ddml::mdl_glm wrapper for stats::glm().

ddml 0.1.0

• Initial CRAN submission.

ddml documentation built on Oct. 3, 2024, 1:13 a.m.