A comparison of agent-based models and the parametric g-formula for causal inference.

This github repository contains SAS code for creating simulated longitudinal data with treatment-confounder feedback under a range of true treatment effects. The repository also contains code for estimating the total effect of treatment using the parametric g-formula and agent-based models under a variety of assumptions about the underlying causal structure when this is unknown. This code was originally published as part of the supplementary material for analyses for Murray et al, 2017.