TorchVinecop.loglik

TorchVinecop.loglik(u, *, x=None)

Total log-likelihood sum(log c(u)) of the vine at u.

Parameters:

uarray, shape (n, d), dtype float

Pseudo-observations in [0, 1]^d.

xarray, shape (n, p), or None, optional

External covariates for a non-simplified vine, else None.

Returns:

array, shape (), dtype float

The summed log-density (a differentiable scalar under autograd, e.g. PyTorch); the per-observation vine density is floored at 1e-20 before the log.