Multidimensional minimization with V-shaped minimum

[Stephan Lorenzen]

Dear list,

I try to minimize a function which has a "peak minimum", meaning it is
V-shaped with discontinuous first derivative at the minimum. So, the
first derivative never comes close to zero. During the minimization I
see that functions f and df are called with minimal values; however the
minimization cancels after the first iteration giving me GSL_ENOPROG
(iteration is not making progress towards solution). What is bad:The
values in gsl_multimin_fdfminimizer_x(m) are NOT those giving the
minimum, but the starting values!

How can I get the parameters of the minimum (lowest point traversed
during iterations) if gsl_multimin_fdfminimizer_x(m) gives wrong values?

Thanks,

Stephan

--
Dr. Stephan Lorenzen
Mühlenredder 35
21493 Schwarzenbek
Tel.: 0177 872 8817