Struct NewtonRaphsonSolver 

pub struct NewtonRaphsonSolver { /* private fields */ }

Newton-Raphson solver for systems of nonlinear equations

This solver can handle:

• Square systems (equations == variables)
• Under-constrained systems (equations < variables) - uses least squares
• Over-constrained systems (equations > variables) - uses least squares

The solver uses adaptive damping and regularization for numerical stability.

Implementations

impl NewtonRaphsonSolver

pub fn new(compiled: CompiledSystem) -> Self

Create a new Newton-Raphson solver with adaptive parameters

Parameters are automatically adjusted based on system type:

• Over-constrained systems get more iterations, relaxed tolerance, conservative damping
• Normal systems use standard parameters for fast convergence

Arguments

• compiled - Compiled system of equations (each should evaluate to 0 at solution)

pub fn new_with_mode( compiled: CompiledSystem, mode: Mode, ) -> Result<Self, SolverError>

Create a new solver with an explicit execution mode.

pub fn new_with_variables( compiled: CompiledSystem, variables: &[&str], ) -> Result<Self, SolverError>

Create a solver while specifying which variables to solve for.

Variables not listed here are treated as fixed parameters and must be provided in the initial guess.

pub fn new_with_variables_and_mode( compiled: CompiledSystem, variables: &[&str], mode: Mode, ) -> Result<Self, SolverError>

Create a solver with explicit variables and execution mode.

pub fn with_mode(self, mode: Mode) -> Result<Self, SolverError>

Update the execution mode for this solver.

pub fn mode(&self) -> &Mode

Returns the current execution mode.

pub fn try_with_equation_traces( self, traces: Vec<Option<EquationTrace>>, ) -> Result<Self, SolverError>

Attach metadata describing the origin of each equation in the system

pub fn with_equation_traces(self, traces: Vec<Option<EquationTrace>>) -> Self

Attach metadata describing the origin of each equation in the system

pub fn trace_for_equation( &self, equation_index: usize, ) -> Option<&EquationTrace>

Fetch trace metadata for a specific equation, if available

pub fn with_max_iterations(self, max_iterations: usize) -> Self

Override the maximum number of iterations

pub fn with_tolerance(self, tolerance: f64) -> Self

Override the convergence tolerance

pub fn with_damping(self, damping_factor: f64) -> Self

Override the damping factor (clamped to [0.1, 1.0] for stability)

pub fn with_regularization(self, regularization: f64) -> Self

Override the base regularization parameter used when the system is ill-conditioned

pub fn solve( &self, initial_guess: HashMap<String, f64>, ) -> Result<Solution, SolverError>

Solve the system using standard Newton-Raphson method

pub fn solve_with_line_search( &self, initial_guess: HashMap<String, f64>, ) -> Result<Solution, SolverError>

Solve the system using Newton-Raphson with line search

Line search helps improve convergence robustness by automatically finding good step sizes, especially useful for difficult systems.