VineDensity.cdf

VineDensity.cdf(X, N=10000, random_state=None)

Evaluates the joint CDF at the given samples.

Returns \(\hat F(\mathbf{x}) = \hat C(\hat F_1(x_1), \ldots, \hat F_d(x_d))\) by applying the marginal CDFs to obtain pseudo-observations and evaluating the fitted copula CDF via the backend’s quasi-Monte-Carlo routine.

Parameters:

Xndarray, shape (n_samples, n_features), dtype float, or DataFrame

Test samples.

Nint, default=10000

Number of quasi-random points used for the Monte-Carlo integration. Larger N gives more accurate CDF values at the cost of more compute.

random_stateint, RandomState instance or None, default=None

Resolved fresh via sklearn.utils.check_random_state for this call. None reuses the RNG resolved at fit time.

Returns:

ndarray, shape (n_samples,), dtype float

Joint CDF values in \([0, 1]\). Values are stochastic across calls with different random_state; use a larger N if the MC noise floor is significant for your application.