Nash variant of Marquardt nonlinear least squares solution via qr linear solver.

Description

Given a nonlinear model expressed as an expression of the form lhs ~ formula_for_rhs and a start vector where parameters used in the model formula are named, attempts to find the minimum of the residual sum of squares using the Nash variant (Nash, 1979) of the Marquardt algorithm, where the linear sub-problem is solved by a qr method.

Usage

1
2
   wrapnls(formula, start, trace=FALSE, data, lower=-Inf, upper=Inf, 
           control=list(), ...)

Arguments

formula

This is a modeling formula of the form (as in nls) lhsvar ~ rhsexpression for example, y ~ b1/(1+b2*exp(-b3*tt)) You may also give this as a string.

start

A named parameter vector. For our example, we could use start=c(b1=1, b2=2.345, b3=0.123)

trace

Logical TRUE if we want intermediate progress to be reported. Default is FALSE.

data

A data frame containing the data of the variables in the formula. This data may, however, be supplied directly in the parent frame.

lower

Lower bounds on the parameters. If a single number, this will be applied to all parameters. Default -Inf.

upper

Upper bounds on the parameters. If a single number, this will be applied to all parameters. Default Inf.

control

A list of controls for the algorithm. These are as for nlxb().

...

Any data needed for computation of the residual vector from the expression rhsexpression - lhsvar. Note that this is the negative of the usual residual, but the sum of squares is the same.

Details

wrapnls first attempts to solve the nonlinear sum of squares problem by using nlsmnq, then takes the parameters from that method to call nls.

Value

An object of type nls.

Note

Special notes, if any, will appear here.

Author(s)

John C Nash <nashjc@uottawa.ca>

See Also

Function nls(), packages optim and optimx.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
cat("See examples in nlmrt-package.Rd\n")

## Not run: 
cat("kvanderpoel.R test\n")
# require(nlmrt)
x<-c(1,3,5,7)
y<-c(37.98,11.68,3.65,3.93)
pks28<-data.frame(x=x,y=y)
fit0<-try(nls(y~(a+b*exp(1)^(-c*x)), data=pks28, start=c(a=0,b=1,c=1), 
          trace=TRUE))
print(fit0)
cat("\n\n")
fit1<-nlxb(y~(a+b*exp(-c*x)), data=pks28, start=c(a=0,b=1,c=1), trace = TRUE)
print(fit1) 
cat("\n\nor better\n")
fit2<-wrapnls(y~(a+b*exp(-c*x)), data=pks28, start=c(a=0,b=1,c=1), 
              lower=-Inf, upper=Inf, trace = TRUE)


## End(Not run)