for testing and deploying your application
for finding and fixing issues
for empowering human code reviews
# Author: Simon Blanke
# Email: [email protected]
# License: MIT License
import numpy as np
from ..base_optimizer import BaseOptimizer
from ...search import Search
from numpy.random import normal, laplace, logistic, gumbel
dist_dict = {
"normal": normal,
"laplace": laplace,
"logistic": logistic,
"gumbel": gumbel,
}
class HillClimbingOptimizer(BaseOptimizer, Search):
def __init__(
self, search_space, epsilon=0.05, distribution="normal", n_neighbours=1,
):
super().__init__(search_space)
self.epsilon = epsilon
self.distribution = dist_dict[distribution]
self.n_neighbours = n_neighbours
def _move_climb(self, pos, epsilon_mod=1):
sigma = self.space_dim * self.epsilon * epsilon_mod
pos_normal = self.distribution(pos, sigma, pos.shape)
pos_new_int = np.rint(pos_normal)
n_zeros = [0] * len(self.space_dim)
pos = np.clip(pos_new_int, n_zeros, self.space_dim)
self.pos_new = pos.astype(int)
return self.pos_new
def iterate(self):
return self._move_climb(self.pos_current)
SimonBlanke /
Gradient-Free-Optimizers
