ghssN1g: Internal Function
Description
It provides the log-likelihood, gradient and Hessian matrix for penalized maximum likelihood optimization
of all parameters but the dispersion and association coefficients, when using the bivariate normal copula with normal and gamma margins.