1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
|
\name{Rvmminb}
\alias{Rvmminb}
\encoding{UTF-8}
\title{Variable metric nonlinear function minimization with bounds constraints}
\description{A bounds-constarined R implementation of a variable metric method for minimization
of nonlinear functions subject to bounds (box) constraints and masks
(fixed parameters).
See manual Rvmmin.Rd for more details and examples.
}
\usage{
Rvmminb(par, fn, gr, lower, upper, bdmsk, control = list(), \dots)
}
\arguments{
\item{par}{A numeric vector of starting estimates.}
\item{fn}{A function that returns the value of the objective at the
supplied set of parameters \code{par} using auxiliary data in \dots.
The first argument of \code{fn} must be \code{par}. }
\item{gr}{A function that returns the gradient of the objective at the
supplied set of parameters \code{par} using auxiliary data in \dots.
The first argument of \code{fn} must be \code{par}. This function
returns the gradient as a numeric vector.
Note that a gradient function MUST be provided. See the manual for
\code{Rvmmin}, which is the usual way \code{Rvmminb} is called. The
user must take responsibility for errors if \code{Rvmminb} is called
directly.
}
\item{lower}{A vector of lower bounds on the parameters.}
\item{upper}{A vector of upper bounds on the parameters.}
\item{bdmsk}{An indicator vector, having 1 for each parameter that is "free" or
unconstrained, and 0 for any parameter that is fixed or MASKED for the
duration of the optimization.}
\item{control}{
An optional list of control settings. See the manual Rvmmin.Rd for
details.
}
\item{\dots}{Further arguments to be passed to \code{fn}.}
}
\details{
This routine is intended to be called from \code{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.
Note that bounds checking, if it is carried out, is done by \code{Rvmmin}.
Functions \code{fn} must return a numeric value.
}
\value{
A list with components:
\item{par}{The best set of parameters found.}
\item{value}{The value of the objective at the best set of parameters found.}
\item{counts}{A vector of two integers giving the number of function and gradient evaluations.}
\item{convergence}{An integer indicating the situation on termination of the function. \code{0}
indicates that the method believes it has succeeded. Other values:
\describe{
\item{\code{1}}{indicates that the iteration limit \code{maxit}
had been reached.}
\item{\code{20}}{indicates that the initial set of parameters is inadmissible, that is,
that the function cannot be computed or returns an infinite, NULL, or NA value.}
\item{\code{21}}{indicates that an intermediate set of parameters is inadmissible.}
}
}
\item{message}{A description of the situation on termination of the function.}
\item{bdmsk}{Returned index describing the status of bounds and masks at the
proposed solution. Parameters for which bdmsk are 1 are unconstrained
or "free", those with bdmsk 0 are masked i.e., fixed. For historical
reasons, we indicate a parameter is at a lower bound using -3
or upper bound using -1.}
}
\seealso{\code{\link{optim}}}
\examples{
## See Rvmmin.Rd
}
\keyword{nonlinear}
\keyword{optimize}
|
