C++ log-likelihood for mixed beta interval regression (Laplace) — .brsmm_loglik_laplace

Computes the marginal log-likelihood for a random-intercept beta interval model using a Laplace approximation of group-specific integrals over Gaussian random effects.

Usage

.brsmm_loglik_laplace_cpp(
  param,
  X,
  Z,
  y_left,
  y_right,
  yt,
  delta,
  group,
  link_mu_code,
  link_phi_code,
  repar
)

Arguments

param: Parameter vector: beta (p), gamma (q), log_sigma_b (1).
X: Mean-model design matrix.
Z: Precision-model design matrix.
y_left: Left interval endpoints.
y_right: Right interval endpoints.
yt: Midpoint responses.
delta: Censoring indicator vector.
group: Integer group index (1..G), one per observation.
link_mu_code: Integer code for mean link.
link_phi_code: Integer code for precision link.
repar: Integer beta reparameterization code.

Value

Scalar marginal log-likelihood.