Struct Model 

pub struct Model { /* private fields */ }

A linear programming model built using the algebraic modeling API.

Implementations

impl Model

pub fn new(name: &str) -> Self

Create a new, empty model with the given name.

pub fn set_timeout(&mut self, secs: f64)

Set a timeout for QP solve operations.

pub fn try_set_timeout(&mut self, secs: f64) -> Result<&mut Self, ModelError>

Set a timeout for solve operations, returning an error for invalid input.

pub fn set_threads(&mut self, n: usize) -> &mut Self

並列 thread 上限を設定する。0 は 1 に補正、default は SolverOptions の 1。 LP / QP / 非凸 multistart すべてに影響する共通設定。

pub fn set_use_ruiz_scaling(&mut self, flag: bool)

Ruiz equilibration スケーリングの有効/無効を設定する（デフォルト: true）

pub fn set_tolerance(&mut self, tol: Tolerance) -> &mut Self

精度プリセットを設定する。

pub fn set_presolve(&mut self, flag: bool) -> &mut Self

Presolve の有効/無効を設定する（デフォルト: true）。

pub fn add_var(&mut self, name: &str, lb: f64, ub: f64) -> Variable

Add a decision variable to the model.

Records an error (deferred to solve()) if lb > ub or either bound is NaN. Use try_add_var to get an immediate Result.

pub fn try_add_var( &mut self, name: &str, lb: f64, ub: f64, ) -> Result<Variable, ModelError>

Add a decision variable, returning an error for invalid bounds.

pub fn add_int_var(&mut self, name: &str, lb: f64, ub: f64) -> Variable

Add an integer decision variable (must take integral values within [lb, ub]).

Records an error (deferred to solve()) if lb > ub or either bound is NaN. Use try_add_int_var to get an immediate Result.

Presence of any integer/binary variable routes solve() through the MILP/MIQP branch-and-bound solver instead of the continuous LP/QP path.

pub fn try_add_int_var( &mut self, name: &str, lb: f64, ub: f64, ) -> Result<Variable, ModelError>

Add an integer decision variable, returning an error for invalid bounds.

pub fn add_binary_var(&mut self, name: &str) -> Variable

Add a binary decision variable (integer, fixed to the {0, 1} box).

pub fn var_name(&self, var: Variable) -> &str

Return the name of a variable as given to add_var.

Panics

Panics if var.index is out of range. Passing a Variable from a different model is unchecked and may index into unrelated data. Use try_var_name for a checked variant.

pub fn try_var_name(&self, var: Variable) -> Result<&str, ModelError>

Return the name of a variable, returning an error instead of panicking.

Returns Err if:

• var was created by a different model.
• var.index is out of range for this model’s variable list.

pub fn add_constraint(&mut self, c: Constraint) -> &mut Self

Add a constraint to the model.

pub fn minimize(&mut self, obj: impl Into<QuadExpr>) -> &mut Self

Set the objective to minimize the given expression.

Accepts any linear Expression or quadratic QuadExpr (e.g. x * x, x * y, x.pow2()). When quadratic terms are present the Q matrix is built automatically using the 1/2 xᵀQx convention and the model is routed through the QP solver.

pub fn maximize(&mut self, obj: impl Into<QuadExpr>) -> &mut Self

Set the objective to maximize the given expression.

See minimize for details on quadratic support. For maximization the Q matrix is negated internally (Q → -Q).

pub fn set_obj_offset(&mut self, offset: f64) -> &mut Self

Set the quadratic objective Q matrix for QP problems.

Convention (“1/2あり”): the objective is min 1/2 x^T Q x + c^T x 目的関数の定数オフセットを設定する。 objective_value = (1/2 x^T Q x +) c^T x + offset として最終結果に加算される。

pub fn solve(&mut self) -> Result<ModelResult, ModelError>

Solve the model and return the result.

Errors

• ModelError::NoObjective if minimize or maximize was not called.
• ModelError::SolveError if the solver returns Infeasible or Unbounded.