Open dslramsey opened 3 years ago
This specific situation should be fixed in 7fc9ca5e534ccb1d423a56edaf9d2ea4f8626b45. I'll leave the issue open because I haven't addressed the situation where an entire year of observations are NA, which gets complicated for this model.
Sorry for the radio silence. Just got around to testing the latest changes and the issue has been now resolved for me.
Missing values in y for multmixOpen() models causes ranef() to fail. A reproducible example follows.
The issue appears to be in the call to dmultinom() where NA's are trapped for capture probability but not for y.