R help - Apr 2010 - optimisation, pseudo maximum likehood and family exponential in R

Hello,

By fixing log_vraisemblance, gradient and hessian function linked to a
family density like exp(theta.xi), I''m looking for some efficients
estimators by PML.

So I''ve seen optim,nlminb, et maxLik procedure.

But I ''m not sure that the heteroscedacity of my estimators are
considered.

Does anyone has been deal with kind of problem?

Thanks for your help.

 

 


	[[alternative HTML version deleted]]