Dean Zenner

Hi aoguedao, interestingly I ran into the exact same problem today. You have to downgrade scikit-learn to Version 1.3.0. I also opened an issue today but closed it. There you...

Hey @Joao-Dionisio for sure! I have a matrix M of shape (output features, parameters). Let's say the amount of parameters is N. I want to maximize over a binary array...

```python # this is my custom function def collinearity(matrix): matrix_mult = torch.matmul(matrix.T, matrix) all_eigenvals = torch.real(torch.linalg.eigvals(matrix_mult)) smallest_eigenval = torch.min(all_eigenvals) return 1 / torch.sqrt(smallest_eigenval) class CollinearityConshdlr(Conshdlr): def __init__(self, S, threshold, xvars):...