Causal Inference using Multivariate Generalized Propensity Score

Methods for estimating weights and generalized propensity score for multiple continuous exposures via the generalized propensity score described in Williams, J.R, and Cresi, C.M (2020) . Weights are constructed assuming an underlying multivariate normal density for the marginal and conditional distribution of exposures given a set of confounders. These weights can then be used to estimate dose-response curves or surfaces. This method achieves balance across all exposure dimension rather than along a single dimension.