PyHeuristics is a Metaheuristics toolkit, which facilitates the use of various metaheuristic algorithms, such as Simulated Annealing or Ant Colony Optimization among others.
TODO
Using the various abstractions of PyHeuristics, it is really easy to solve an optimization problem.
Using the various abstractions of PyHeuristics, it is really easy to solve an optimization problem. For example, we will show how to find the maximum of this expression using the Simulated Annealing algorithm.
First of all import SimulatedAnnealing and CoolingSchedule. CoolingSchedule
is a class that regulated the temperature of the annealing process.
from pyheuristics.trajectory.annealing.cooling_schedules import GeometricCoolingSchedule
from pyheuristics.trajectory.annealing.simulated_annealing import SimulatedAnnealingThen, implement a class of the problem that must implement the fn and
the move methods.
class MaximizeSimpleFunction(SimulatedAnnealing):
def fn(self, sol: Any) -> float:
binary_to_decimal = int(sol, 2)
return -(
binary_to_decimal**3
- 60 * (binary_to_decimal**2)
+ 900 * binary_to_decimal
+ 100
)
def move(self) -> Any:
bit_to_flip = random.randint(0, 4)
_list = list(self.sol)
if self.sol[bit_to_flip] == "1":
_list[bit_to_flip] = "0"
else:
_list[bit_to_flip] = "1"
return "".join(_list)Finally, run the algorithm to search for the maximum value.
cs = GeometricCoolingSchedule(initial_temp=25000.0, final_temp=2.5, alpha=0.9)
optim_problem = MaximizeSimpleFunction(init_sol="10011", cooling_schedule=cs)
execution_result = optim_problem.run()
print(execution_result.sol) # "01010"
print(execution_result.cost) # 4100
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent