Detailed Description

Solver Policy 1 (Diagonal): Cholesky decomposition using W.

This solver is used when hessian_block_size == 1. It computes the diagonal of the negative Hessian of the log-likelihood, takes its square root, and forms the system matrix B = I + W_r * Sigma * W_r for Cholesky factorization.

The solver is the fastest option but only valid when the Hessian is truly diagonal (no cross-terms between latent variables).

Note: This solver corresponds to solver == 1 in the original code.

Definition at line 467 of file laplace_marginal_density_estimator.hpp.

#include <laplace_marginal_density_estimator.hpp>

Public Member Functions
template<typename NewtonStateT , typename CovarMat >
	CholeskyWSolverDiag (const NewtonStateT &state, const CovarMat &covariance)

template<typename NewtonStateT , typename LLFun , typename LLTupleArgs , typename CovarMat >
void	solve_step (NewtonStateT &state, const LLFun &ll_fun, const LLTupleArgs &ll_args, const CovarMat &covariance, int, std::ostream *msgs)
	Perform one Newton step using diagonal Hessian solver.

double	compute_log_determinant () const
	Compute log determinant of B from Cholesky factor.

template<typename NewtonStateT >
auto	build_result (NewtonStateT &state, double log_det)
	Build the final result structure.

Public Attributes
Eigen::VectorXd	W_r_diag
	Square root of diagonal Hessian: W_r[j] = sqrt(W[j])

Eigen::VectorXd	W_diag
	Diagonal Hessian values from the likelihood.

Eigen::LLT< Eigen::MatrixXd >	llt_B
	Cholesky factorization of B = I + W_r * Sigma * W_r.

The documentation for this struct was generated from the following file:

stan/math/mix/functor/laplace_marginal_density_estimator.hpp

Detailed Description

Public Member Functions

Public Attributes