def search(self, n=10, retries=20):
"""
Attempts to find the optimal solution to the n-Queens problem using local search
where the maximum number of retries is given. Each retry has a maximum of 'n'
search steps.
:param n: The maximum number of search steps for each retry.
:param retries: The number of retries.
:return: The optimum found by the local search algorithm.
"""
for retry in range(retries):
curr_optimum = Board(self.size)
for step in range(n):
x = self.search_step(curr_optimum)
if x is None:
break
curr_optimum = x
if curr_optimum.cost() < self.optimal_cost:
self.optimum = curr_optimum
self.optimal_cost = curr_optimum.cost()
if self.optimal_cost == 0:
return self.optimum
return self.optimum
class LocalSearch:
def __init__(self, size: int):
self.size = size
self.optimum = Board(size)
self.optimal_cost = self.optimum.cost()
def search_step(self, optimum: Board):
"""
Step searches the new optimal neighbour in a greedy manner. The optimal neighbour is returned
if there exists an improvement. Otherwise, None is returned.
:param optimum: The board from which the neighbourhood will be checked.
:return: The improved Board or None if there does not exist any.
"""
optimal_cost = optimum.cost()
curr_optimum = optimum
curr_cost = optimal_cost
for neighbour in curr_optimum.neighbourhood():
if curr_optimum is None or curr_cost > neighbour.cost():
curr_optimum = neighbour
curr_cost = neighbour.cost()
if curr_cost < optimal_cost:
return curr_optimum
else:
return None
def search(self, n=10, retries=20):
"""
Attempts to find the optimal solution to the n-Queens problem using local search
where the maximum number of retries is given. Each retry has a maximum of 'n'
search steps.
:param n: The maximum number of search steps for each retry.
:param retries: The number of retries.
:return: The optimum found by the local search algorithm.
"""
for retry in range(retries):
curr_optimum = Board(self.size)
for step in range(n):
x = self.search_step(curr_optimum)
if x is None:
break
curr_optimum = x
if curr_optimum.cost() < self.optimal_cost:
self.optimum = curr_optimum
self.optimal_cost = curr_optimum.cost()
if self.optimal_cost == 0:
return self.optimum
return self.optimum
def search_step(self, optimum: Board):
"""
Step searches the new optimal neighbour in a greedy manner. The optimal neighbour is returned
if there exists an improvement. Otherwise, None is returned.
:param optimum: The board from which the neighbourhood will be checked.
:return: The improved Board or None if there does not exist any.
"""
optimal_cost = optimum.cost()
curr_optimum = optimum
curr_cost = optimal_cost
for neighbour in curr_optimum.neighbourhood():
if curr_optimum is None or curr_cost > neighbour.cost():
curr_optimum = neighbour
curr_cost = neighbour.cost()
if curr_cost < optimal_cost:
return curr_optimum
else:
return None
