Comprehensive Correlation Model Inference — correlation_comprehensive

Compute all three types of inference (Hessian, Bootstrap, Profile) for DCC or CGARCH correlation parameters.

Usage

correlation_comprehensive_inference(
  model_type = c("dcc", "cgarch", "adcc"),
  residuals,
  weights,
  Qbar = NULL,
  mle_params = NULL,
  distribution = "mvn",
  n_boot = 200,
  boot_method = "residual",
  n_profile_points = 50,
  conf_level = 0.95,
  verbose = TRUE,
  seed = NULL
)

Arguments

model_type: "dcc" or "cgarch"
residuals: T x k matrix of residuals
weights: T-vector of observation weights
Qbar: k x k unconditional correlation matrix (optional)
mle_params: MLE estimates (optional, computed if NULL)
distribution: "mvn" or "mvt"
n_boot: Number of bootstrap replications
boot_method: Bootstrap method
n_profile_points: Points for profile likelihood
conf_level: Confidence level
verbose: Print progress
seed: Random seed

Value

List with all inference results