On hierarchical loglinear models in capture–recapture studies

Hierarchical loglinear models are widely used in capture–recapture studies. It is important to implement these models so that a full model selection procedure can be carried out. An algorithm used to count the number of monotone boolean functions is adopted to generate all the monotone boolean functions, which in turn is used to generate all coefficient matrices of hierarchical loglinear models. The proposed methods are implemented in an R package. Two real examples are analyzed for illustration.