Releases: benkeser/drtmle
Releases · benkeser/drtmle
v1.1.0
drtmle v1.0.5
- Added
nSL
option for fitting and averaging multiple super learners as part
of the estimation procedure. - Added
adapt_g
option for outcome-adaptive propensity score fitting.
drtmle v1.0.4
- Resolves issues arising from
returnModels
option when users input nuisance parameters - Add option to bypass
future
calls for easier debugging - Fixes to bugs in standard TMLE implementation -- namely, more robust fluctuations and corrected variance estimators
Version 1.0.3
Version 1.0.3 on CRAN
drtmle V 1.0.2
February 5, 2018
- Version 1.0.2 released on GitHub and CRAN
- Replace foreach parallelization with future
- More robust super learner methods included
- Fix test to pass build with long doubles removed
- Return estimated influence functions with drtmle fit for power users
- Minor documentation corrections and updates
drtmle v1.0.0
Version 1.0.0 released on CRAN.