GOGARCH Model Diagnostics

Comprehensive diagnostics for GOGARCH model fit.

Usage

gogarch_diagnostics(gogarch_result, residuals, verbose = TRUE)

Arguments

gogarch_result

Result from estimate_garch_weighted_gogarch or estimate_gogarch_improved

residuals

Original T x k residual matrix

verbose

Print diagnostic report (default: TRUE)

Value

List with:

ica_quality

Independence score, negentropy, reconstruction error

component_diagnostics

Per-component: persistence, Ljung-Box p-value

mixing_matrix

Orthogonality error, condition number

covariance_fit

RMSE, correlation vs sample covariance

Details

Key quality thresholds:

• Independence score > 0.8

• Reconstruction error < 1\

• Mixing matrix condition number < 100

• Component persistence < 1 (stationarity)

See also

estimate_garch_weighted_gogarch, improved_ica_decomposition