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
|
\name{grfwd}
\alias{grfwd}
\alias{optsp}
\title{Forward difference numerical gradient approximation.}
\description{
\code{grfwd} computes the forward difference approximation to the gradient of
user function \code{userfn}.
}
\usage{
grfwd(par, userfn, fbase=NULL, env=optsp, ...)
}
\arguments{
\item{par}{
parameters to the user objective function userfn
}
\item{userfn}{
User-supplied objective function
}
\item{fbase}{
The value of the function at the parameters, else NULL. This is to save
recomputing the function at this point.
}
\item{env}{
Environment for scratchpad items (like \code{deps} for approximation
control in this routine). Default \code{optsp}.
}
\item{\dots}{
optional arguments passed to the objective function.
}
}
\details{
\tabular{ll}{
Package: \tab grfwd\cr
Depends: \tab R (>= 2.6.1)\cr
License: \tab GPL Version 2.\cr
}
}
\value{
\code{grfwd} returns a single vector object \code{df} which approximates the
gradient of userfn at the parameters par. The approximation is controlled by a
global value \code{optderiveps} that is set when the package is attached.
}
\author{
John C. Nash
}
\examples{
cat("Example of use of grfwd\n")
myfn<-function(xx, shift=100){
ii<-1:length(xx)
result<-shift+sum(xx^ii)
}
xx<-c(1,2,3,4)
ii<-1:length(xx)
print(xx)
gn<-grfwd(xx,myfn, shift=0)
print(gn)
ga<-ii*xx^(ii-1)
cat("compare to\n")
print(ga)
}
\keyword{optimize}
|
