Result

OptimizeResult

givp.OptimizeResult dataclass

Container for optimization output, modeled after scipy.optimize.OptimizeResult.

Attributes:

Name	Type	Description
x	NDArray[float64]	Best solution vector found, in the user's variable space.
fun	float	Objective value at x, in the user's original sign (i.e., already restored when direction='maximize').
nit	int	Number of GRASP outer iterations executed.
nfev	int	Number of objective function evaluations.
success	bool	Whether at least one feasible solution was produced.
message	str	Human-readable termination reason.
direction	str	'minimize' or 'maximize'.
meta	AlgorithmMeta	Additional algorithm-specific information (cache stats, etc.).

to_dict

to_dict() -> dict[str, Any]

Serialize to a JSON-safe dict with typed, schema-validated fields.

All values are Python primitives — no numpy types. The termination field is a TerminationReason enum value (closed set), which prevents prompt-injection via free-form message strings when passed to LLM agents.

---

TerminationReason

givp.TerminationReason

Bases: str, Enum

Closed set of termination reasons — safe to pass to LLM agents.

from_message classmethod

from_message(message: str) -> TerminationReason

Map a free-form termination message to the nearest enum value.