run_newton_loop()

template<typename SolverPolicy , typename NewtonStateT , typename OptionsT , typename LLFunT , typename LLTupleArgsT , typename CovarMatT , typename UpdateFun >

auto stan::math::internal::run_newton_loop ( SolverPolicy &  solver, NewtonStateT &  state, const OptionsT &  options, Eigen::Index &  step_iter, const LLFunT &  ll_fun, const LLTupleArgsT &  ll_args, const CovarMatT &  covariance, UpdateFun &&  update_fun, std::ostream *  msgs  )

inline

Run a Newton loop with a solver policy, updating the shared state.

This helper centralizes the iteration/line-search logic shared across solver policies while preserving the step counter and fallback behavior.

Template Parameters

SolverPolicy	Either Solver 1, 2, or 3 policy struct
NewtonStateT	Type holding state with solver
OptionsT	struct of options for laplace
LLFunT	Callable accepting (ThetaVec, LLTupleArgs...)
LLTupleArgsT	tuple of additional args for LLFunT
CovarMatT	Type of the covariance matrix

Parameters

[in,out]	solver	The solver policy instance
[in,out]	state	Shared Newton optimization state
[in]	options	Solver options
[in,out]	step_iter	Current iteration number (modified)
[in]	ll_fun	Log-likelihood functor
[in]	ll_args	Additional arguments for the likelihood
[in]	covariance	Prior covariance matrix Sigma
[in]	update_fun	Callable to update proposal state in line search
[in,out]	msgs	Output stream for diagnostic messages (may be nullptr)

Returns

laplace_density_estimates built from solver policy

Stop when objective change is small, or when a rejected Wolfe step fails to improve; finish_update then exits the Newton loop.

Definition at line 927 of file laplace_marginal_density_estimator.hpp.