R help - Mar 2008 - Error en nlm(logdgenexpn, p = c(vmomest[[1]], vmomest[[2]]), x = x.genexp, : valor no finito provisto por 'nlm'

Dear useRs,

I am analysing the behaviour of MLE for the two parameters of a kind
of exponential distribution, leaving as initial values the estimators
moments produced by the variation coefficient.

 I do using simulations, giving them an accountant, r. But running my
codes remains a problem with the nlm function. To review details
wearing

On one of the lines put status

cond04<-is.finite(mle.nlm$minimum)

        {
         .....

An Introduction to the Interactive Debugging Tools in R, Roger D.
Peng, UCLA Department of Statistics, August 28, 2002




The error message is delivered:

Error en nlm(logdgenexpn, p = c(vmomest[[1]], vmomest[[2]]), x = x.genexp,  :
        valor no finito provisto por 'nlm'
Adem?s: Hubo 50 o  m?s avisos (use warnings() para ver los primeros 50)


One possibility is to re-generate the data set that produces such
error, but not know how to do it.

Someone knows how to keep in an object  the parameters of the
generator which produces such shows.

l can send codes for details.

Thank you for your attention.

Bernardo.
University of Concepci?n