perry (version 0.1.1)

repRS: (Repeated) random splitting for fitted models

Description

Estimate the prediction error of a fitted model via (repeated) random splitting (also known as random subsampling or Monte Carlo cross-validation). This works for any model for which a perry method is available.

Usage

repRS(object, m, R = 1, grouping = NULL, splits = NULL,
    ...)

Arguments

object

the fitted model for which to estimate the prediction error.

m

an integer giving the number of observations or groups of observations to be used as test data.

R

an integer giving the number of random data splits.

grouping

a factor specifying groups of observations. If supplied, the data are split according to the groups rather than individual observations such that all observations within a group belong either to the training or test data.

splits

an object of class "randomSplits" (as returned by randomSplits) or a control object of class "splitControl" (see splitControl<

...

additional arguments to be passed down to perry.

Value

• An object of class "perry" as returned by perry.

See Also

perry, repCV, bootPE

Examples

# load data and fit an LS regression model
data("mtcars")
fit <- lm(mpg ~ wt + cyl, data=mtcars)

# perform random splitting
repRS(fit, m = 6, R = 10, seed = 1234)