Hi there, I'm new to tidymodels and am interested in implementing a common model for ecological analysis (Occupancy models for those familiar, paper unfortunately paywalled). This model has two main 'parameters' to be estimated; probability of 'occupancy' and probability of 'detection' given occupancy. Each can be estimated with additional covariates. Data used to fit the model are detection histories (a series of binary survey outcomes within a year), and the probability of recording a given detection history is a function of 'occupancy' and 'detection.' These parameters are jointly estimated in a maximum likelihood framework, which is where I'm having trouble - I understand how to implement a single response variable, or two independent response variables, but not how to implement two joint parameters like these. Does tidymodels have the capacity to handle this?
Thanks!
I dont know this scientific arena, but there appears to be a modern R package designed for such; CRAN - Package unmarked
This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.
If you have a query related to it or one of the replies, start a new topic and refer back with a link.