Re: COPULA.RPF—Copula estimation
Posted: Sun Jun 23, 2024 9:59 am
Hi,
How to do a full log likelihood estimation in RATS?
From Trivedi and Zimmer the loglikelihood is
LN(b1,b2,c) = L1,N (b1,b2) + L2,N (b1,b2,c).
where is the second part on the right hand of the equation involves the dependence parameter.
We are interested in b1 and b2, which are the parameters in the marginal functions. We would like to take the log likelihood of the whole function, and thus get the parameters b1 and b2.
Can you upload an example of that?
The example you post just estimate the dependence parameter. It is enlightening and helpful.
However, if we have a parametric model for the marginals, and we would like to re-estimate the parameters in the marginals together with the dependence parameter in the copula in a full log-likelihood functional, is there a way to do that in RATS?
Thank you!
How to do a full log likelihood estimation in RATS?
From Trivedi and Zimmer the loglikelihood is
LN(b1,b2,c) = L1,N (b1,b2) + L2,N (b1,b2,c).
where is the second part on the right hand of the equation involves the dependence parameter.
We are interested in b1 and b2, which are the parameters in the marginal functions. We would like to take the log likelihood of the whole function, and thus get the parameters b1 and b2.
Can you upload an example of that?
The example you post just estimate the dependence parameter. It is enlightening and helpful.
However, if we have a parametric model for the marginals, and we would like to re-estimate the parameters in the marginals together with the dependence parameter in the copula in a full log-likelihood functional, is there a way to do that in RATS?
Thank you!