Rearrangement (version 2.1)

simconboot: Simultaneous Confidence Interval Estimation using Bootstrap

Description

simconboot obtains a simultaneous confidence interval for a function. It estimates the lower and upper endpoint functions of the interval by bootstrap.

Usage

simconboot(x,y,estimator,formula,B=200,alpha=0.05,sampsize=length(x), seed=8,colInt=c(5:39)/2,...)

Arguments

x

a numerical vector of x values

y

a numerical vector of y values

estimator

estimator to be used in regression

formula

formula to be used in the estimator

B

an integer with the number of bootstrap repetitions

alpha

a real number between 0 and 1 reflecting the desired confidence level

sampsize

an integer with the sample size of each bootstrap repetition

seed

if desired, seed to be set for the random number generator

colInt

the points to be evaluated when ploting

...

further arguments to be passed to the estimator

Value

An object of class conint with the following elements:

x

the original x data

y

the original y data

sortedx

the original x data, sorted with repeated elements removed

Lower

the lower endpoint function. Represented as a vector of values corresponding to sortedx

Upper

the upper endpoint function. Represented as a vector of values corresponding to sortedx

cef

the corresponding estimates

Details

estimator can be any of a set of standard regression models, most commonly lm or rq (from package quantreg) for global estimators and the built-in functions lclm, lplm, lcrq2, lprq2 for local estimators.

Note: formula=0 for all the local estimators.

See Also

rconint

Examples

data(GrowthChart)
attach(GrowthChart)

nage <- 2 * pi * (age - min(age)) / (max(age) - min(age))
nages <- unique(sort(nage))
formula <- height~I(sin(nage))+I(cos(nage))+I(sin(2*nage))+I(cos(2*nage))+
            I(sin(3*nage))+I(cos(3*nage))+I(sin(4*nage))+I(cos(4*nage))
j <- simconboot(nage,height,lm,formula)
plot(j, border=NA, col='darkgray',xlab = 'Age (years)',ylab = 'Height (cms)',xaxt = "n")
axis(1, at = seq(-2*pi*min(age)/(max(age)-min(age)), 2*pi+1,
        by=5*2*pi/(max(age)-min(age))), label = seq(0, max(age)+1, by=5))
points(nage,height)
lines(nages, j$cef, lty=2, col='green')

detach(GrowthChart)