FitControlsConfig struct
Configuration options for initializing a FitControlsVinecop object.
This struct provides a flexible way to initialize a FitControlsVinecop object. Each field is optional, and default values are applied if the field is not set.
Public variables
- std::optional<std::vector<BicopFamily>> family_set
- The set of bicop families to consider. Defaults to all families.
- std::optional<std::string> parametric_method
- Method for parametric estimation (e.g., "mle"). Default: "mle".
- std::optional<std::string> nonparametric_method
- std::optional<double> nonparametric_mult
-
std::optional<size_
t> nonparametric_grid_size - Grid size for nonparametric estimation. Default: 30.
- std::optional<std::string> selection_criterion
- Criterion for model selection (e.g., "aic"). Default: "aic".
- std::optional<Eigen::VectorXd> weights
- Observation weights. Default: an empty Eigen::VectorXd.
- std::optional<double> psi0
- Threshold for the psi0 parameter. Default: 0.9.
- std::optional<bool> preselect_families
- std::optional<bool> allow_rotations
- Whether to allow rotations for the families. Default: true.
-
std::optional<size_
t> num_threads - Number of threads to use during fitting. Default: 1.
-
std::optional<size_
t> trunc_lvl - Truncation level for truncated vines. Default: no truncation.
- std::optional<std::string> tree_criterion
- The criterion for selecting the maximum spanning tree. Default: "tau".
- std::optional<TreeCriterionFunction> tree_criterion_function
- std::optional<double> threshold
- Threshold for thresholded vines. Default: 0.
- std::optional<bool> select_threshold
- Whether to select the truncation level automatically. Default: false.
- std::optional<bool> select_trunc_lvl
- Whether to select the threshold automatically. Default: false.
- std::optional<bool> select_families
- Whether to select the families automatically. Default: true.
- std::optional<bool> show_trace
- Whether to show a trace of the building progress. Default: false.
- std::optional<std::string> tree_algorithm
- The algorithm for building the maximum spanning tree. Default: "mst_prim".
- std::optional<std::vector<int>> seeds
- A vector of random seeds for the random number generator.
-
std::optional<std::vector<size_
t>> conditioning_set
Variable documentation
std::optional<std::string> vinecopulib::
Method for nonparametric estimation (e.g., "constant"). Default: "constant".
std::optional<double> vinecopulib::
Method for nonparametric estimation (e.g., "constant"). Default: "constant".
std::optional<bool> vinecopulib::
Whether to preselect families based on preliminary criteria. Default: true.
std::optional<TreeCriterionFunction> vinecopulib::
A custom edge-weight function for the spanning tree. Required when tree_criterion is set to "custom", and rejected otherwise. It is always called on the thread that starts the fit, so it need not be thread safe.
std::optional<std::vector<size_
Conditioning set for conditioning-aware structure selection: 1-based variable indices to place at the tail of the vine order (so they can be conditioned on with Vinecop::). Default: empty (unconditional selection).