Profile Likelihood Confidence Intervals

Compute profile likelihood CIs for correlation parameters. Supports DCC and CGARCH models.

Usage

correlation_profile_likelihood_ci(
  model_type = c("dcc", "cgarch"),
  residuals,
  weights,
  Qbar,
  mle_params,
  mle_nll = NULL,
  conf_level = 0.95,
  param = c("both", "alpha", "beta"),
  n_points = 50,
  distribution = "mvn",
  verbose = TRUE
)

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

mle_params

MLE estimates

mle_nll

NLL at MLE

conf_level

Confidence level (default 0.95)

param

Which parameter: "alpha", "beta", or "both"

n_points

Number of profile points

distribution

Distribution type

verbose

Print progress

Value

List with profile likelihood results