Hi all,
I am trying to perform an ordinal regression using proportional odds
model.
First I would like to test whether the assumption of proportional odds is
supported by the data.
I got the following error message when I try to fit one of the relevant
models. Can anyone give
some suggestions on how to fix this. Thank you.
Hanna
fit2 <- vglm(PLANHORIZON ~ factor(TotalCompanion) +
factor(NumTrips)+factor(Numnight)+factor(FMWEBSITE)+factor(OTHERWEB)+factor(BROCHU)
+ +factor(TRAAGENT) + factor(WOM) + factor(AAA) + factor(GUIDEBOOK) +
factor(ADVERTISE) + factor(NEWSPAPER) + factor(INFOOTHER) +
+ ExpTotal + factor(AGE) + factor(CHILD) + factor(INCOME), +
family=cumulative(parallel=F), data=two)
Error in if (any(smallmu <- (mu * (1 - mu) < double.eps))) { :
missing value where TRUE/FALSE needed
In addition: Warning messages:
1: In Deviance.categorical.data.vgam(mu = mu, y = y, w = w, residuals residuals,
:
fitted values close to 0 or 1
2: In Deviance.categorical.data.vgam(mu = mu, y = y, w = w, residuals residuals,
:
fitted values close to 0 or 1
[[alternative HTML version deleted]]