Spaces:
Sleeping
Sleeping
| import torch | |
| import numpy as np | |
| from botorch.test_functions import Ackley | |
| device = torch.device("cpu") | |
| dtype = torch.double | |
| def Ackley500D(individuals): | |
| # assert torch.is_tensor(individuals) and individuals.size(1) == 10, "Input must be an n-by-10 PyTorch tensor." | |
| ############################################################################# | |
| ############################################################################# | |
| # Set function here: | |
| dimm = individuals.shape[1] | |
| fun = Ackley(dim=dimm, negate=True) | |
| fun.bounds[0, :].fill_(-5) | |
| fun.bounds[1, :].fill_(10) | |
| dim = fun.dim | |
| lb, ub = fun.bounds | |
| ############################################################################# | |
| ############################################################################# | |
| n = individuals.size(0) | |
| fx = fun(individuals) | |
| fx = fx.reshape((n, 1)) | |
| ############################################################################# | |
| ## Constraints | |
| gx1 = torch.sum(individuals,1) - 1000 # sigma(x) <= 0 | |
| gx1 = gx1.reshape((n, 1)) | |
| gx2 = torch.norm(individuals, p=2, dim=1) - 115 # norm_2(x) -3 <= 0 | |
| gx2 = gx2.reshape((n, 1)) | |
| gx = torch.cat((gx1, gx2), 1) | |
| ############################################################################# | |
| return gx, fx | |
| def Ackley500D_Scaling(X): | |
| # print(X.shape) | |
| # assert torch.is_tensor(X) and X.size(1) == 10, "Input must be an n-by-10 PyTorch tensor." | |
| X_scaled = X*15.0 - 5.0 | |
| return X_scaled | |