Hi, there, Is there a R routine which can fit multinomial logistic regression for nominal outcomes? Not the multinom() of log-linear model, neither the polr() for ordinal outcomes. Thanks. Jun Han