Combined Conditional-Poisson/Poisson Hessian, Standalone (C++)

Computes the Hessian matrix of the log-likelihood of the combined KK matched-pair conditional-Poisson (conditional-Binomial) plus reservoir marginal-Poisson model documented in full at fast_cpoisson_combined_with_var_cpp, at arbitrary caller-supplied params_r (not necessarily the MLE). Internally reuses the same score-and-information computation as get_cpoisson_combined_score_cpp (a single shared routine computes both at once) and returns the negative of the resulting information matrix, i.e. the actual Hessian of the log-likelihood. Exported standalone — independent of any optimizer run — for direct numerical diagnostics at a specific parameter value.

Usage

get_cpoisson_combined_hessian_cpp(
  yT_v_r,
  n_k_v_r,
  X_diff_v_r,
  y_r_r,
  w_r_r,
  X_r_r,
  params_r
)

Arguments

yT_v_r

Treated-subject outcome count per matched pair.

n_k_v_r

Total (treated + control) outcome count per matched pair.

X_diff_v_r

Covariate differences (treated minus control) between the members of each matched pair.

y_r_r

Reservoir (unmatched) subjects' outcomes.

w_r_r

Reservoir subjects' treatment indicators.

X_r_r

Reservoir subjects' covariates.

params_r

A numeric vector of model parameters at which to evaluate the Hessian.

Value

The Hessian matrix of the log-likelihood (the negative of the information matrix) at params_r.

See also

get_cpoisson_combined_score_cpp for the corresponding gradient at the same point; fast_cpoisson_combined_with_var_cpp for the full model documentation.