Steve_Friedman at nps.gov
2010-Feb-10 16:32 UTC
[R] mlogit: Error reported using sample dataset
I've been working on a multinomial logit model, trying to predict vegetation types as a function of total phosphorus. Previous responses to my postings have pointed me to the mlogit package. I'm now trying to work examples and my data using this package. data("Fishing", package = "mlogit") Fish <- mlogit.data(Fishing, varying = c(4:11), shape = "wide", choice "mode") both of those lines work, as expected. However; summary(mlogit(mode ~ pr + ca -1, data = Fish)) Call: Error in deparse(object at call) : trying to get slot "call" from an object (class "mlogit") that is not an S4 object Using my data I've done the following: xx <- fwc # copies and protects original data.frame xx$VegType <- as.factor(xx$VegType) xx.dat <- mlogit.data(xx, varying = "NULL", choice = "VegType", shape "wide") xx.mlogit <- mlogit(VegType ~ TP, data = xx.dat,, reflevel = "1") Error in relevel.factor(attr(mf, "index")[["alt"]], reflevel) : 'ref' must be an existing level Can someone explain why these errors are occuring and what to do about them? Thanks Using R 2.10.1, Windows XP Steve Friedman Ph. D. Spatial Statistical Analyst Everglades and Dry Tortugas National Park 950 N Krome Ave (3rd Floor) Homestead, Florida 33034 Steve_Friedman at nps.gov Office (305) 224 - 4282 Fax (305) 224 - 4147