An R implementation of a variable metric method for minimization of unconstrained nonlinear functions.
See the manual Rvmmin.Rd for details.
Rvmminu(par, fn, gr, control = list(), ...)A list with components:
The best set of parameters found.
The value of the objective at the best set of parameters found.
A vector of two integers giving the number of function and gradient evaluations.
An integer indicating the situation on termination of the function. 0
indicates that the method believes it has succeeded. Other values:
1indicates that the iteration limit maxit
had been reached.
20indicates that the initial set of parameters is inadmissible, that is, that the function cannot be computed or returns an infinite, NULL, or NA value.
21indicates that an intermediate set of parameters is inadmissible.
A description of the situation on termination of the function.
A numeric vector of starting estimates.
A function that returns the value of the objective at the
supplied set of parameters par using auxiliary data in ....
The first argument of fn must be par.
A function that returns the gradient of the objective at the
supplied set of parameters par using auxiliary data in ....
The first argument of fn must be par. This function
returns the gradient as a numeric vector.
Note that a gradient function MUST be provided. See the manual for
Rvmmin, which is the usual way Rvmminu is called. The
user must take responsibility for errors if Rvmminu is called
directly.
An optional list of control settings. See the manual Rvmmin.Rd for details. Some control elements apply only when parameters are bounds constrained and are not used in this function.
Further arguments to be passed to fn.
This routine is intended to be called from Rvmmin, which will, if
necessary, supply a gradient approximation. However, some users will want
to avoid the extra overhead, in which case it is important to provide an
appropriate and high-accuracy gradient routine.
Functions fn must return a numeric value.