Finite mixture models (FMMs) can be used in settings where some unmeasured classification separates the observed data into groups with different exposure/outcome relationships. One familiar example of this is a zero-inflated model, where some observat...