Dear all,
I'd like to ask if it would be a good idea to add an extra argument to
'optim' (with default value NULL) to give a initial Hessian for the
BFGS algorithm. This could be handy in some cases where this initial
approximation is available and thus it isn't required 'optim' to
compute it again.
Thanks for any comments,
Dimitris
----
Dimitris Rizopoulos
Ph.D. Student
Biostatistical Centre
School of Public Health
Catholic University of Leuven
Address: Kapucijnenvoer 35, Leuven, Belgium
Tel: +32/16/336899
Fax: +32/16/337015
Web: http://www.med.kuleuven.ac.be/biostat/
http://www.student.kuleuven.ac.be/~m0390867/dimitris.htm