Skip to contents

Compute DCC(1,1) Hessian with Full Diagnostics

dcc11_hessian.Rd

Core function for Hessian-based inference. Returns Hessian, variance-covariance matrix, standard errors, and diagnostic information (eigenvalues, condition number) useful for detecting the "flat beta problem".

Usage

dcc11_hessian(
  params,
  std_resid,
  weights,
  Qbar,
  distribution = "mvn",
  use_reparam = FALSE,
  hessian_method = "numerical",
  eps = 1e-05
)

Arguments

params

MLE parameter estimates c(alpha, beta) or c(alpha, beta, shape)

std_resid

T x k matrix of standardized residuals

weights

T-vector of observation weights

Qbar

k x k unconditional correlation matrix

distribution

"mvn" or "mvt"

use_reparam

Logical: parameters in reparameterized (psi, phi) space?

hessian_method

"numerical" (default) or "analytical"

eps

Step size for numerical differentiation

Value

List with:

hessian

Hessian matrix of NLL

info

Observed information matrix (= Hessian for NLL)

vcov

Variance-covariance matrix (inverse of info)

se

Standard errors

eigenvalues

Eigenvalues of Hessian

eigenvectors

Eigenvectors of Hessian

condition_number

Condition number of Hessian

params

Parameters

param_names

Para,eter names

method

"hessian"