usl (version 1.2.1)

confint: Confidence Intervals for USL model parameters

Description

Estimate confidence intervals for one or more parameters in a USL model.

Usage

## S3 method for class 'USL':
confint(object, parm, level = 0.95, type =
"norm")

Arguments

object

A USL object.

parm

A specification of which parameters are to be given confidence intervals, either a vector of numbers or a vector of names. If missing, all parameters are considered.

level

The confidence level required.

type

A character string representing the type of interval required. The value must be one of "norm", "basic", "stud", "perc" or "bca".

Value

• A matrix (or vector) with columns giving lower and upper confidence limits for each parameter. These will be labelled as (1-level)/2 and 1 - (1-level)/2 in % (by default 2.5% and 97.5%).

Details

Confidence intervals are estimated by using bootstrap resampling (see boot). The bootstrap generates a series of random selections with replacement for the original data and calculates the model parameters sigma and kappa for every selection. This set of parameters is used to estimate confidence intervals.

The parameter "type" determines the type of interval that is calculated. The values "norm", "basic", "stud", "perc" or "bca" are currently allowed. See boot.ci for details.

The bootstrap is computed once in the usl function so calling confint multiple times for a specific USL object will return identical results. Creating multiple USL objects for a given set of input values is almost certainly going to produce different confidence intervals since random numbers are used to bootstrap.

Calculating confidence intervals for a small number of observations is unreliable. The function will print warning or error messages if the calculated intervals are dubious or the estimation is not possible.

See Also

usl, boot, boot.ci

Examples

require(usl)

data(specsdm91)

## Create USL model
usl.model <- usl(throughput ~ load, specsdm91)

## Print normal confidence intervals
confint(usl.model, type = "norm")