Re: [R] maxNR in maxLik package never stops

Hi, sure, maxNR uses 'iterlim' (or was it 'maxIter'?) where you can specify the max # of iterations. The default was 100 iterations, AFAIK. There was a bug in one of the previous versions which lead to infinite loops of constrained maximization. Do you use constrained or unconstrained? and yes

Re: [R] Robust SE & Heteroskedasticity-consistent estimation

) > } > > > So S <-B(resMaxlik)*M(resMaxlik)*B(resMaxlik) directly. It's ok. > > But I call sandwich function like this : > sandwich(resMaxlik,meat.=M,bread.=B) > It returns a error message : "Erreur dans UseMethod("estfun") : > pas de méthode pour 'e

Re: [R] Interpreting coefficient in selection and outcome Heckman models in sampleSelection

Hi Mark, why do you need that? If your task is to estimate how much your y changes if x change, why not use simple OLS? (Well, right, you should be able to use sampleSelection as well). It shouldn't probably be hard to compute it -- it is just OLS marginal effect + som kind of derivative of Inve