In lnalborczyk/ESTER: Efficient Sequential Testing with Evidence Ratios

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "README-"
)

Efficient Sequential Testing with Evidence Ratios

The ESTER package implements sequential testing based on evidence ratios computed from the weights of a set of models. These weights correspond to Akaike weights when based on either the Akaike Information Criterion (AIC) or the Bayesian Information Criterion (BIC), and to pseudo-BMA weights when computed from the Widely Applicable Information Criterion (WAIC).

Installation

You can install the latest published version from CRAN using:

install.packages("ESTER")

Or the development version (recommended) from Github with:

if (!require("devtools") ) install.packages("devtools")
devtools::install_github("lnalborczyk/ESTER", dependencies = TRUE)

How to use the package ?

Model comparison

The ictab function takes as input a named list of models to be compared, and returns a dataframe with the given information criterion and the weight of each model.

library(ESTER)
data(mtcars)

mod1 <- lm(mpg ~ cyl, mtcars)
mod2 <- lm(mpg ~ cyl + vs, mtcars)
mod3 <- lm(mpg ~ cyl * vs, mtcars)

mods <- list(mod1 = mod1, mod2 = mod2, mod3 = mod3)

ictab(mods, aic)

Sequential testing

You can study the evolution of sequential ERs using the seqtest function.

data(mtcars)

mod1 <- lm(mpg ~ cyl, mtcars)
mod2 <- lm(mpg ~ cyl + disp, mtcars)

seqtest(ic = aic, mod1, mod2, nmin = 10)

More detailed information can be found in the main vignette, available online here, or by typing vignette("ESTER") in the console.

lnalborczyk/ESTER documentation built on May 21, 2019, 7:36 a.m.