outfit {MQMF}

outfit tidy print of output from optim, nlminb, or nlm

Description

outfit takes in the output list from either optim, nlminb, or nlm and prints it more tidily to the console, In the case of nlm it also prints the conclusion regarding the solution. It might be more effective to implement an S3 method.

Usage

outfit(inopt, backtran = TRUE, digits = 5, title = "", parnames = "")

Arguments

inopt

the list object output by nlm, nlminb, or optim

backtran

a logical default = TRUE If TRUE it assumes that the parameters have been log-transformed for stability and need back-transforming

digits

the number of digits to round the backtransformed parameters. defaults to 5.

title

character string used to label the output if desired, default = empty character string

parnames

default="" which means the estimated parameters will merely be numbered. If a vector of names is given then this will be used instead, at least, for nlm and optim.

Value

nothing but it does print the list to the console tidily

Examples

 x <- 1:10  # generate power function data from c(2,2) + random
 y <- c(2.07,8.2,19.28,40.4,37.8,64.68,100.2,129.11,151.77,218.94)
 alldat <- cbind(x=x,y=y)
 pow <- function(pars,x) return(pars[1] * x ^ pars[2])
 ssq <- function(pars,indat) {
    return(sum((indat[,"y"] - pow(pars,indat[,"x"]))^2))
 }  # fit a power curve using normal random errors
 pars <- c(2,2)
 best <- nlm(f=ssq,p=pars,typsize=magnitude(pars),indat=alldat)
 outfit(best,backtran=FALSE) #a=1.3134, b=2.2029 ssq=571.5804