predict.logistf with insufficient number of factor levels

[agleiss]

Hi Georg,
[MK said it is worth an issue... hope it is]
I fitted a model with logistf including a character column (not as factor but with, say, 4 different values). Trying to get predictions with predict on new data that only include one of the categories of the character column gives the error message

Fehler in contrasts<-(*tmp*, value = contr.funs[1 + isOF[nn]]) :
Kontraste können nur auf Faktoren mit 2 oder mehr Stufen angewendet werden

Using new data that contain more than one different but not all categories gives the error message

Fehler in object$coefficients %*% t(X) : nicht passende Argumente

Using additional dummy rows representing all possible categories gives the requested predictions.

Thank you,
Andreas