SimonBlanke /
Gradient-Free-Optimizers
| Total Complexity | 1 |
| Total Lines | 25 |
| Duplicated Lines | 0 % |
| Changes | 0 | ||
| 1 | import numpy as np |
||
| 2 | from gradient_free_optimizers import StochasticHillClimbingOptimizer |
||
| 3 | |||
| 4 | |||
| 5 | def sphere_function(para): |
||
| 6 | x = para["x"] |
||
| 7 | y = para["y"] |
||
| 8 | |||
| 9 | return -(x * x + y * y) |
||
| 10 | |||
| 11 | |||
| 12 | search_space = { |
||
| 13 | "x": np.arange(-10, 10, 0.1), |
||
| 14 | "y": np.arange(-10, 10, 0.1), |
||
| 15 | } |
||
| 16 | |||
| 17 | opt = StochasticHillClimbingOptimizer( |
||
| 18 | search_space, |
||
| 19 | epsilon=0.01, |
||
| 20 | n_neighbours=5, |
||
| 21 | distribution="laplace", |
||
| 22 | p_accept=0.05, |
||
| 23 | ) |
||
| 24 | opt.search(sphere_function, n_iter=10000) |
||
| 25 |
