Open jimianelli opened 9 years ago
The other way to achieve this would be to introduce the sex-ratio as a time-varying random variable. This way, a beta (or Dirichlet) could be used to place an informative prior on what this sex-ratio should be. This would essentially be the same number of parameters/latent states being estimated, but the sex-ratio would be well-defined.
would use control-file feature linked to number of sexes...