fit_mst: Fit MST mixture model

In carter-allen/BayesMSN: Bayesian Multivariate Skew Normal Mixture Models

Description

This function fits the MST mixture model without imputation

Usage

fit_mst(Y, X, W, K, nu = 3, nsim = 1000, burn = 0)

Arguments

Y	matrix of non-missing outcome data
X	matrix of predictors used in the MSN model component
W	matrix of predictors used in the multinomial logit clustering component
K	number of clusters to fit. Use WAIC to choose best K.
nu	degrees of freedom parameter for skew-t model.
nsim	number of MCMC iterations to perform
burn	number of initial MCMC iterations to discard. The function will return nsim - burn total posterior samples.

Examples

data(example1_data)
fit1 <- fit_mst(Y = example1_data$Y,
                X = example1_data$X,
                W = example1_data$W,
                K = 3,
                nsim = 10,
                burn = 0)

carter-allen/BayesMSN documentation built on Aug. 22, 2020, 8:27 a.m.