AugBalWeight 0.1.0
- Initial CRAN release of
AugBalWeight.
- Implements augmented balancing weights as linear regression based on
Bruns-Smith, Dukes, Feller, and Ogburn (2026, JRSSB, DOI:
10.1093/jrsssb/qkaf019).
- Supports
double_ridge (l2 balancing),
double_lasso (l-infinity balancing), and general augmented
outcome models.
- Includes hyperparameter cross-validation for outcome models,
covariate balance, and Riesz loss.
- Provides
aug_bal_ate and aug_bal_att
wrappers for causal inference with treatment effect estimation.
- Features balance diagnostic plotting, robust EIF standard errors,
and counterfactual prediction methods.