Source code for metaheuristic_designer.strategies.hill_climb

"""
Hill Climbing strategy (single-solution, greedy local improvement).
"""

from __future__ import annotations
from typing import Optional
from ..initializer import Initializer
from ..survivor_selection_base import SurvivorSelection
from ..search_strategy import SearchStrategy
from ..operator import Operator
from ..survivor_selection import create_survivor_selection
from ..utils import check_random_state, RNGLike




class HillClimb(SearchStrategy):
    """
    Hill Climbing algorithm.

    A single solution is perturbed each iteration.  If the new
    solution is better, it replaces the current one.  By default,
    the survivor selection is set to one-to-one competition
    (``"hill_climb"`` in the survivor registry).

    Parameters
    ----------
    initializer : Initializer
        Population initializer (typically creates a single individual).
    operator : Operator, optional
        Perturbation operator.  Defaults to :class:`NullOperator`.
    survivor_sel : SurvivorSelection, optional
        Survivor selection method; defaults to ``"hill_climb"``.
    params : dict, optional
        Additional parameters stored as schedulable values.
    name : str, optional
        Display name (default ``"HillClimb"``).
    random_state : RNGLike, optional
        Random number generator.
    **kwargs
        Forwarded to :class:`SearchStrategy`.
    """

    def __init__(
        self,
        initializer: Initializer,
        operator: Optional[Operator] = None,
        survivor_sel: Optional[SurvivorSelection] = None,
        params: Optional[dict] = None,
        name: str = "HillClimb",
        random_state: Optional[RNGLike] = None,
        **kwargs,
    ):

        if survivor_sel is None:
            survivor_sel = create_survivor_selection("hill_climb")

        super().__init__(initializer, operator=operator, survivor_sel=survivor_sel, params=params, name=name, random_state=random_state, **kwargs)