g-computation

Epidemiology analysis package

Taking causal inference to the extreme!

An R-Package to Visualize the (Causal) Effect of a Continuous Variable on a Time-To-Event Outcome

Targeted maximum likelihood estimation (TMLE) enables the integration of machine learning approaches in comparative effectiveness studies. It is a doubly robust method, making use of both the outcome model and propensity score model to generate an unbiased estimate as long as at least one of the models is correctly specified.

The R package trajmsm is based on the paper Marginal Structural Models with Latent Class Growth Analysis of Treatment Trajectories: https://doi.org/10.48550/arXiv.2105.12720.

Course repository for BIOST 578: Causal Inference for Biomedical Studies (Spring 2024) with Dr. Ting Ye

Functional Machine Learning in R

This paper uses g-computation to estimate the effect of long-term exposure to air-pollution on mortality

Main repository for the 2nd paper of my PhD.

R code for the analyses conducted in Friedrich, S & Friede, T (2020). Causal inference methods for small non-randomized studies: Methods and an application in COVID-19. Submitted to Contemporary Clinical Trials.

Estimated marginal means, contrasts, and small-sample mixed-model inference for Python — a native port of R's emmeans and pbkrtest.

Causal inference: the effect of smoking cessation on weight gain (NHEFS) via g-computation, IPW and doubly-robust estimation, with balance and sensitivity checks

Deterministic selective-labels harness: measures PD models on the declined population real data never sees, and prices it in profit. Plants ground truth; the operating frontier is a 25-seed distribution, not a point. sklearn-only, byte-reproducible.

Counterfactual life-course editing for longitudinal social processes in R

Code for Harnois-Leblanc, Rifas-Shiman, Switkowski et al. (2025) Estimating sex-specific population-level effects AJE

Lumbar Fusion Digital Twin (LF-DT): calibrated multi-outcome prediction with g-computation counterfactual re-simulation of operative decisions, benchmarked against doubly robust AIPW.