Vinecop

class Vinecop(*args, **kwargs)

A class for vine copula models.

A vine copula model is characterized by its structure (see RVineStructure objects) and the pair-copulas (see Bicop objects).

Attributes:

dim: Returns ——- int the dimension of the vine (number of variables).
families: Returns ——- list of list of BicopFamily the families of all pair copulas, indexed as [tree][edge].
matrix: Returns ——- ndarray, shape (n, m), dtype int the R-vine structure as a square matrix (cf.
nobs: Returns ——- int the number of observations used to fit the vine (0 until fit() has been called).
npars: Returns ——- float the total number of parameters across all pair copulas.
order: Returns ——- list of int the natural variable order in the first tree.
pair_copulas: Returns ——- list of list of Bicop all pair copulas, indexed as [tree][edge].
parameters: Returns ——- list of list of ndarray, shape (n, m), dtype float the parameter matrices of all pair copulas, indexed as [tree][edge].
rotations: Returns ——- list of list of int the rotations of all pair copulas, indexed as [tree][edge].
structure: Returns ——- RVineStructure the underlying R-vine structure.
taus: Returns ——- list of list of float Kendall’s tau for every pair copula, indexed as [tree][edge].
threshold: Returns ——- float the absolute-dependence threshold used during structure selection (0 if thresholding was disabled).
trunc_lvl: Returns ——- int the truncation level (number of fitted trees; pair copulas in trees above this level are forced to independence).
var_types: Returns ——- list of str the variable types of each variable (each "c" for continuous or "d" for discrete).

Methods

`aic`(self[, u, num_threads])	Evaluates the Akaike information criterion (AIC).
`bic`(self[, u, num_threads])	Evaluates the Bayesian information criterion (BIC).
`cdf`(self, u[, N, num_threads, seeds])	Evaluates the copula distribution.
`fit`(self, data[, controls, num_threads])	Fits the parameters of a pre-specified vine copula model.
`format`(self[, trees])	Summarizes the model into a string (can be used for printing).
`from_data`(data[, structure, matrix, ...])	Factory function to create a Vinecop from data.
`from_dimension`(d)	Instantiates a D-vine with all pair-copulas set to independence.
`from_file`(filename[, check])	Instantiates from a JSON or CBOR file.
`from_json`(json[, check])	Instantiates from a nlohmann::json object.
`from_structure`([structure, matrix, ...])	Factory function to create a Vinecop using either a structure or a matrix.
`get_family`(self, tree, edge)	Gets the family of the pair copula at the given (tree, edge).
`get_pair_copula`(self, tree, edge)	Gets the pair copula at the given (tree, edge) position.
`get_parameters`(self, tree, edge)	Gets the parameter matrix of the pair copula at the given (tree, edge).
`get_rotation`(self, tree, edge)	Gets the rotation (degrees) of the pair copula at the given (tree, edge).
`get_struct_array`(self[, natural_order])	Parameters:
`get_tau`(self, tree, edge)	Gets Kendall's tau of the pair copula at the given (tree, edge).
`gradient`(self, u[, step_wise, num_threads])	Evaluates the gradient of the average log-likelihood.
`hessian`(self, u[, step_wise, num_threads])	Evaluates the averaged Hessian of the log-likelihood.
`hessian_full`(self, u[, step_wise, num_threads])	Evaluates the hessian per observation.
`inverse_rosenblatt`(self, u[, num_threads])	Evaluates the inverse Rosenblatt transform.
`loglik`(self[, u, num_threads])	Evaluates the log-likelihood.
`mbicv`(self[, u, psi0, num_threads])	Evaluates the modified Bayesian information criterion for vines (mBICV).
`pdf`(self, u[, num_threads])	Evaluates the copula density.
`pdf_full`(self, u[, num_threads, keep_all])	Evaluates the vine copula density and, optionally, the per-edge densities and h-functions.
`plot`(self[, tree, add_edge_labels, layout, ...])	Generates a plot for the Vinecop object.
`rosenblatt`(self, u[, num_threads, ...])	Evaluates the Rosenblatt transform for a vine copula model.
`scores`(self, u[, step_wise, num_threads])	Evaluates the score function.
`scores_cov`(self, u[, step_wise, num_threads])	Computes the covariance matrix of scores.
`scores_full`(self, u[, step_wise, ...])	Evaluates the score function together with the per-edge derivative caches.
`select`(self, data[, controls])	In other words, `select()` behaves differently depending on its current truncation level and the truncation level specified in the controls, respectively called `trunc_lvl` and `controls.trunc_lvl` in what follows.
`simulate`(self, n[, qrng, num_threads, seeds])	Simulates from a vine copula model, see `inverse_rosenblatt()`.
`to_file`(self, filename)	Writes the copula object into a JSON or CBOR file.
`to_json`(self)	Converts the copula into a nlohmann::json object.
`truncate`(self, trunc_lvl)	Truncates the vine copula model.

Attributes

`dim`	Returns:
`families`	Returns:
`matrix`	Returns:
`nobs`	Returns:
`npars`	Returns:
`order`	Returns:
`pair_copulas`	Returns:
`parameters`	Returns:
`rotations`	Returns:
`structure`	Returns:
`taus`	Returns:
`threshold`	Returns:
`trunc_lvl`	Returns:
`var_types`	Returns:

Methods

`__init__`	Default constructor for the `Vinecop` class.
`aic`	Evaluates the Akaike information criterion (AIC).
`bic`	Evaluates the Bayesian information criterion (BIC).
`cdf`	Evaluates the copula distribution.
`fit`	Fits the parameters of a pre-specified vine copula model.
`format`	Summarizes the model into a string (can be used for printing).
`from_data`	Factory function to create a Vinecop from data.
`from_dimension`	Instantiates a D-vine with all pair-copulas set to independence.
`from_file`	Instantiates from a JSON or CBOR file.
`from_json`	Instantiates from a nlohmann::json object.
`from_structure`	Factory function to create a Vinecop using either a structure or a matrix.
`get_family`	Gets the family of the pair copula at the given (tree, edge).
`get_pair_copula`	Gets the pair copula at the given (tree, edge) position.
`get_parameters`	Gets the parameter matrix of the pair copula at the given (tree, edge).
`get_rotation`	Gets the rotation (degrees) of the pair copula at the given (tree, edge).
`get_struct_array`	Parameters:
`get_tau`	Gets Kendall's tau of the pair copula at the given (tree, edge).
`gradient`	Evaluates the gradient of the average log-likelihood.
`hessian`	Evaluates the averaged Hessian of the log-likelihood.
`hessian_full`	Evaluates the hessian per observation.
`inverse_rosenblatt`	Evaluates the inverse Rosenblatt transform.
`loglik`	Evaluates the log-likelihood.
`mbicv`	Evaluates the modified Bayesian information criterion for vines (mBICV).
`pdf`	Evaluates the copula density.
`pdf_full`	Evaluates the vine copula density and, optionally, the per-edge densities and h-functions.
`plot`	Generates a plot for the Vinecop object.
`rosenblatt`	Evaluates the Rosenblatt transform for a vine copula model.
`scores`	Evaluates the score function.
`scores_cov`	Computes the covariance matrix of scores.
`scores_full`	Evaluates the score function together with the per-edge derivative caches.
`select`	In other words, `select()` behaves differently depending on its current truncation level and the truncation level specified in the controls, respectively called `trunc_lvl` and `controls.trunc_lvl` in what follows.
`simulate`	Simulates from a vine copula model, see `inverse_rosenblatt()`.
`to_file`	Writes the copula object into a JSON or CBOR file.
`to_json`	Converts the copula into a nlohmann::json object.
`truncate`	Truncates the vine copula model.