Struct OptimProblem 

pub struct OptimProblem<S: Scalar> { /* private fields */ }

Declarative optimization problem.

Build a problem with the fluent API, then call .solve() to automatically dispatch to the best solver, or .solve_with(choice) for explicit control.

Implementations

impl<S: Scalar> OptimProblem<S>

pub fn new(n: usize) -> Self

Create a new problem with n decision variables.

pub fn x0(self, x0: &[S]) -> Self

Set the initial point.

pub fn bounds(self, i: usize, lo_hi: (S, S)) -> Self

Set bounds for variable i.

pub fn all_bounds(self, bounds: &[(S, S)]) -> Self

Set bounds for all variables at once.

pub fn integer_var(self, i: usize) -> Self

Mark variable i as integer-valued.

pub fn binary_var(self, i: usize) -> Self

Mark variable i as binary (0 or 1).

pub fn objective<F: Fn(&[S]) -> S + 'static>(self, f: F) -> Self

Set a scalar objective function to minimize.

pub fn gradient<G: Fn(&[S], &mut [S]) + 'static>(self, g: G) -> Self

Set the gradient for the scalar objective.

pub fn maximize<F: Fn(&[S]) -> S + 'static>(self, f: F) -> Self

Set a scalar objective function to maximize.

Internally converts to minimization by negating the objective and gradient.

pub fn maximize_gradient<G: Fn(&[S], &mut [S]) + 'static>(self, g: G) -> Self

Set the gradient for a maximization objective.

Note: provide the gradient of the ORIGINAL function (not negated). The builder handles negation internally.

pub fn least_squares<R: Fn(&[S], &mut [S]) + 'static>( self, n_residuals: usize, residual: R, ) -> Self

Set a least-squares objective: minimize ||r(x)||^2.

pub fn jacobian<J: Fn(&[S], &mut [S]) + 'static>(self, j: J) -> Self

Set the Jacobian for a least-squares objective.

pub fn linear_objective(self, c: &[S]) -> Self

Set a linear objective: minimize c^T x.

pub fn quadratic_objective_dense(self, h_row_major: &[S], c: &[S]) -> Self

Set a quadratic objective: minimize 1/2 x^T H x + c^T x. h_row_major is the n x n Hessian in row-major order.

pub fn linear_constraint_ineq(self, a: &[S], b: S) -> Self

Add a linear inequality constraint: a^T x <= b.

pub fn linear_constraint_eq(self, a: &[S], b: S) -> Self

Add a linear equality constraint: a^T x = b.

pub fn constraint_ineq<F: Fn(&[S]) -> S + 'static>(self, f: F) -> Self

Add an inequality constraint g(x) <= 0.

pub fn constraint_eq<F: Fn(&[S]) -> S + 'static>(self, f: F) -> Self

Add an equality constraint h(x) = 0.

pub fn constraint_ineq_with_grad<F, G>(self, f: F, g: G) -> Self

where F: Fn(&[S]) -> S + 'static, G: Fn(&[S], &mut [S]) + 'static,

Add an inequality constraint with gradient.

pub fn constraint_eq_with_grad<F, G>(self, f: F, g: G) -> Self

where F: Fn(&[S]) -> S + 'static, G: Fn(&[S], &mut [S]) + 'static,

Add an equality constraint with gradient.

pub fn max_iter(self, n: usize) -> Self

Set maximum iterations.

pub fn gtol(self, tol: S) -> Self

Set gradient tolerance.

pub fn options(self, opts: OptimOptions<S>) -> Self

Override all options.

pub fn aug_lag_options(self, opts: AugLagOptions<S>) -> Self

Set augmented Lagrangian options for constrained problems.

pub fn ctol(self, tol: S) -> Self

Set the constraint tolerance for constrained problems.

pub fn global(self, enabled: bool) -> Self

Enable global optimization (Differential Evolution).

pub fn de_options(self, opts: DEOptions<S>) -> Self

Set Differential Evolution options.

pub fn multi_objective(self, funcs: Vec<ScalarFn<S>>) -> Self

Set multiple objectives for multi-objective optimization (all minimized).

pub fn hint(self, hint: ProblemHint) -> Self

Provide a structural hint about the closure-based objective.

This informs auto solver selection when the objective is a closure that implements a linear or quadratic function but cannot be decomposed into coefficient vectors.

pub fn has_bounds(&self) -> bool

Returns true if any variable has bounds.

pub fn has_constraints(&self) -> bool

Returns true if constraints have been added.

pub fn is_least_squares(&self) -> bool

Returns true if the objective is a least-squares residual.

pub fn is_linear(&self) -> bool

Returns true if the objective is structurally linear (coefficient vector).

pub fn is_quadratic(&self) -> bool

Returns true if the objective is structurally quadratic (Hessian + linear).

pub fn is_hint_linear(&self) -> bool

Returns true if a linear hint is set (closure known to be linear).

pub fn is_hint_quadratic(&self) -> bool

Returns true if a quadratic hint is set (closure known to be quadratic).

pub fn is_multi_objective(&self) -> bool

Returns true if the objective is multi-objective.

pub fn has_linear_constraints(&self) -> bool

Returns true if linear constraints have been added.

pub fn has_integer_vars(&self) -> bool

Returns true if any variable is integer or binary.

pub fn n_eq_constraints(&self) -> usize

Number of equality constraints.

pub fn n_ineq_constraints(&self) -> usize

Number of inequality constraints.

impl<S: Scalar + SimpleEntity + Conjugate<Canonical = S> + ComplexField> OptimProblem<S>

pub fn solve(self) -> Result<OptimResult<S>, OptimError>

Solve using automatic solver selection.

pub fn solve_with( self, choice: SolverChoice, ) -> Result<OptimResult<S>, OptimError>

Solve using an explicitly chosen solver.