StepReg (version 1.5.8)

vote: Vote for all models

Description

Votes for all models across all combinations of strategies and metrics

Usage

vote(x, ...)

Value

A dataframe with column names "model" and combinations of strategy and metric. The first column represents the model formula, and a checkmark indicates that the corresponding model was supported by the given strategy and metric combination. Please note that for the subset strategy, the "vote" will report the single best model across all numbers of variables under Information Criteria (IC). However, this rule should not be applied to Significance Level (SL) because the F/Rao value is only comparable for models with the same number of variables.

Arguments

x

each dataframe from outputlist

...

further parameters

Examples

data(mtcars)
formula <- mpg ~ .
x <- stepwise(formula = formula,
              data = mtcars,
              type = "linear",
              strategy = c("forward","backward","subset"),
              metric = c("AIC","BIC"))
vote(x)