Hi
Apart from nlm() with the stepmax= argument, is there any other >1
dimensional minimisation function where I can limit the step size
dynamically. With the NAG routine E04NBF (I think) I used to bound the
size of the next step using a*arctan(x/a) tricks where a function was
particularly difficult near a boundary and thus 'steer' the result
within
legal limits.
The function I am trying to minimise is (negative) log likelihood for the
GEV but with a *lot* of right-censoring. I am not currently sure whether
it is a numerical problem (it could well) but I would like to explore
any other optimiser if there is one in the many libraries.
\John
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-help-request at
stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
