Learn R Programming

reams (version 0.1)

plursim: Perform a pluralistic model selection simulation study

Description

This function can be used to evaluate model selection performance of AIC, corrected AIC, and resampling-based information criteria. Model selection "success", in the pluralistic sense, is evaluated for each criterion, in each of a set of simulations.

Usage

plursim(n, p.all, p.true, R2, nboot, nsim, resample = "subsampling", nmods = 7)

Arguments

n
sample size.
p.all
maximum model dimension, i.e., number of candidate predictors plus 1.
p.true
true model dimension, i.e., number of predictors with nonzero coefficients plus 1.
R2
coefficient of determination for the true model.
nboot
number of bootstrap samples or subsamples (within each simulation) for the resampling-based information criterion.
nsim
number of simulations.
resample
resampling method: "bootstrap", "subsampling", or an abbreviation of one of these.
nmods
number of "best" models retained as viable models.

Value

  • A list with components AIC, AICc, resampIC, each of which is a table indicating the success rate for the given information criterion in the simulations.