confint.transreg: Confidence intervals for estimated parameters

Description Usage Arguments Value Author(s)

View source: R/transreg.R

Description

Calculates confidence intervals for estimated parameters from a transreg model by inverting the Wald or likelihood ratio tests.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
## S3 method for class 'transreg'
confint(
  treg,
  parm,
  level = 0.95,
  type = "wald",
  lrint = c(0.2, 5),
  extendInt_lower = "downX",
  extendInt_upper = "upX",
  ...
)

Arguments

treg

An object of class transeg.

parm

A parameter name or vector of parameter names. If missing, confidence intervals are calculated for all estimated parameters.

level

The confidence level (1 - alpha).

type

The type of confidence interval. Current options are wald for Wald confidence limits and lr for likelihood ratio confidence limits. The latter are more accurate but more computationally intensive.

lrint

The highest and lowest multiples of the standard error used by uniroot to find likelihood ratio confidence limits.

extendInt_lower

The extendInt argument in uniroot when calculating the lower likelihood ratio confidence limit.

extendInt_upper

The extendInt argument in uniroot when calculating the upper likelihood ratio confidence limit.

...

Additional arguments sent to uniroot.

Value

A data frame with one row for each parameter in parm that contains the lower and upper confidence limits in columns labeled \frac{α}{2} and 1 - \frac{α}{2} expressed as percentiles.

Author(s)

Eben Kenah kenah.1@osu.edu


ekenah/transtat documentation built on Nov. 11, 2021, 3:12 p.m.