utils.integrands package¶
Functions defined on the unit hypercube for the purpose of integration tests
- sanitize_variable(x, device=None)[source]¶
Prepare input variable for a pytorch function: if it is a python numerical variable or a numpy array, then cast it to a tensor if it is is a tensor, make a copy to make sure the function cannot be altered from outside
- Parameters
x (float or np.ndarray or torch.Tensor) –
device (None or torch.device) –
- Returns
- Return type
torch.Tensor
Submodules