I'm trying to estimate a multinomial logit model but in some choices only alternatives from a subset of all possible alternatives can be chosen. At the moment I get around it by creating "dummy" variables to mean the alternative is not available and let it estimate this coefficient as highly negative. Is there a better way to do it?
[[alternative HTML version deleted]] ______________________________________________ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.