from __future__ import annotations
import warnings
from typing import TYPE_CHECKING, Any
import optiland.backend as be
from scipy import optimize
from ..live_plotter import LiveOptimizationPlotter
from .base import OptimizerGeneric
if TYPE_CHECKING:
from ...problem import OptimizationProblem
[docs]
class DualAnnealing(OptimizerGeneric):
"""DualAnnealing is an optimizer that uses the dual annealing algorithm
to find the minimum of an optimization problem.
Args:
problem (OptimizationProblem): The optimization problem to be solved.
Methods:
optimize(maxiter=1000, disp=True): Runs the dual annealing algorithm
to optimize the problem and returns the result.
"""
def __init__(self, problem: OptimizationProblem):
super().__init__(problem)
[docs]
def optimize(
self,
maxiter: int = 1000,
disp: bool = True,
plot: bool = False,
callback: Any = None,
):
"""Runs the dual annealing algorithm to optimize the problem.
Args:
maxiter (int): Maximum number of iterations.
disp (bool): Whether to display the optimization process.
plot: If True, update live plots during optimization.
callback (callable): A callable called after each iteration.
Returns:
result: The result of the optimization.
"""
# Get initial values in backend format
x0_backend = [var.value for var in self.problem.variables]
self._x.append(x0_backend) # Store backend values
# Convert x0 to NumPy for SciPy
x0_numpy = be.to_numpy(x0_backend)
bounds = tuple([var.bounds for var in self.problem.variables])
if any(None in bound for bound in bounds):
raise ValueError("Dual annealing requires all variables have bounds.")
live_plotter: LiveOptimizationPlotter | None = None
if plot:
live_plotter = LiveOptimizationPlotter(self)
live_plotter.initialize()
def _wrapped_callback(*args: Any, **kwargs: Any) -> None:
if callback is not None:
callback(*args, **kwargs)
if live_plotter is not None:
live_plotter.update()
with warnings.catch_warnings():
warnings.simplefilter("ignore", category=RuntimeWarning)
result = optimize.dual_annealing(
self._fun,
bounds=bounds,
maxiter=maxiter,
x0=x0_numpy,
callback=_wrapped_callback,
)
# The last function evaluation is not necessarily the lowest.
# Update all lens variables to their optimized values
for idvar, var in enumerate(self.problem.variables):
var.update(result.x[idvar])
self.problem.update_optics()
if live_plotter is not None:
live_plotter.update()
live_plotter.finalize()
return result