obs_lpmf.stan

Go to the documentation of this file.

 
real obs_lpmf(array[] int obs, vector log_exp_obs, array[] real phi,
              int model_obs) {
  if (model_obs == 0) {
    // Poisson model
    return poisson_log_lpmf(obs | log_exp_obs);
  } else if (model_obs == 1) {
    // NB2 parameterisation: variance = mu + mu^2/phi (constant overdispersion)
    return neg_binomial_2_log_lpmf(obs | log_exp_obs, phi[1]);
  } else {
    // NB1 parameterisation: variance = mu + phi * mu, so phi_nb2 = mu * phi
    return neg_binomial_2_log_lpmf(
      obs | log_exp_obs, exp(log_exp_obs + log(phi[1]))
    );
  }
}
 
real obs_lpmf(int obs, real log_exp_obs, array[] real phi, int model_obs) {
  if (model_obs == 0) {
    // Poisson model
    return poisson_log_lpmf(obs | log_exp_obs);
  } else if (model_obs == 1) {
    // NB2 parameterisation: variance = mu + mu^2/phi (constant overdispersion)
    return neg_binomial_2_log_lpmf(obs | log_exp_obs, phi[1]);
  } else {
    // NB1 parameterisation: variance = mu + phi * mu, so phi_nb2 = mu * phi
    return neg_binomial_2_log_lpmf(
      obs | log_exp_obs, exp(log_exp_obs + log(phi[1]))
    );
  }
}