Software
ipd
Inference on Predicted Data
An open-source R package for downstream modeling of an outcome and its associated features where a potentially sizable portion of the outcome data has been imputed by an AI/ML prediction algorithm. The package implements several recently proposed methods for inference on predicted data (PostPI, PPI, PPI++, PSPA) with a single, user-friendly wrapper function.
CRAN | GitHub | Paper (Bioinformatics)
install.packages("ipd")pgeeVar
Finite-Sample Variance Estimation for Penalized GEE
An open-source R package for fitting penalized generalized estimating equations (PGEE), simulating correlated binary longitudinal data, and computing small-sample covariance estimators for near-separated binary outcomes. The package supports the finite-sample inference workflow developed in my PGEE variance-estimation research.
pak::pak("awanafiaz/pgeeVar")scorcher
Accessible Deep Learning in R
An open-source R package that provides high-level, pipe-compatible functionality for building and fitting deep learning models using the torch library in R. Designed to let clinical and cancer researchers build complex AI/ML models using intuitive, sequential steps rather than hundreds of lines of code.
Supported by the Stuart and Molly Sloan Precision Oncology Institute Technology Dissemination Award at Fred Hutchinson Cancer Center.
GitHub | UW Biostatistics feature
# Install development version
devtools::install_github("jtleek/scorcher")