A reorganization of the call to numDeriv grad() function.

Provides a wrapper for the numDeriv approximation to the gradient of a user supplied objective function userfn.

grnd(par, userfn, ...)

Arguments

par

A vector of parameters to the user-supplied function fn

userfn

A user-supplied function

...

Other data needed to evaluate the user function.

Details

The Richardson method is used in this routine.

Value

grnd returns an approximation to the gradient of the function userfn

Examples

cat("Example of use of grnd\n")
#> Example of use of grnd
require(numDeriv)
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)
#> [1] 1 2 3 4
gn<-grnd(xx,myfn, shift=0)
print(gn)
#> [1]   1   4  27 256
ga<-ii*xx^(ii-1)
cat("compare to\n")
#> compare to
print(ga)
#> [1]   1   4  27 256