[R] Estimating parameters of 3 parameters lognormal distribution

Rolf Turner r.turner at auckland.ac.nz
Sat Jan 18 22:52:13 CET 2014

On 19/01/14 06:20, peter dalgaard wrote:
> Offhand, I'd say that the issue isn't so much the starting values as the risk of hitting the upper bound on gamma. If you get gamma set to min(x) at any point during the iteration, the density at the corresponding point is zero, the log-likelihood is -Inf, and <poof>, the magic smoke escapes.
> Perhaps try min(x)-1e-7 or so?

Yep, that did it.  Thanks Peter.

(Actually I used min(x) - eps where I set eps equal to 
sqrt(.Machine$double.eps) --- which seems to be a kind of standard 
construction for a small-but-not-too-small number.)



More information about the R-help mailing list