Predict fitted MBMM gate probabilities

Predicts the gate probabilities implied by the fitted covariate model before incorporating subject-specific longitudinal mechanistic likelihoods. These are therefore prior latent-class probabilities under the MBMM gate.

Usage

mbmm_predict_gate(fit, newdata)

Arguments

fit

An object of class "mbmm_gate" returned by mbmm_fit_gate.

newdata

A data.frame containing the variables used in fit$formula.

Value

For a binary gate, a numeric vector giving the probability of fit$classes[1]. For a multinomial gate, a matrix with one column per latent class.