confidence: Generic method to compute confidence regions for mmi model...

In JacobBergstedt/mmi: Multiple Model Inference

Generic method to compute confidence regions for mmi model and family objects.

Description

For mmi_lm objects a confidence interval for treatment terms is constructed using the t-distribution assumption for the linear parameter over the standard error. The standard error is estimated using sandwich estimation. For the other models the confidence intervals are computed using the profile likelihood. A tibble will be setup with the information. For a fam object confidence intervals will be computed for each model element using the individual methods, and the tibbles will then be binded together in a map-reduce pattern.

Usage

confidence(object, level, ...)

## S3 method for class 'mmi_model'
confidence(object, level)

## S3 method for class 'mmi_lm'
confidence(object, level)

## S3 method for class 'fam'
confidence(object, level = 0.95)

Arguments

object	Either a mmi model object or a mmi family object.
level	The confidence level.

Methods (by class)

• mmi_model: Confidence method for the superclass mmi_model. Uses confint.

• mmi_lm: Computes the confidence interval using the t-distribution assumption and sets up a tidy tibble with the information. Uses the sandwich estimated standard errors.

• fam: Maps over the family, constructs a confidence tibble for each model object using the corresponding S3 method, and the binds the information together in a tidy way.

JacobBergstedt/mmi documentation built on March 8, 2024, 8:25 a.m.