Plot RFE Performance Profiles

These functions plot the resampling results for the candidate subset sizes evaluated during the recursive feature elimination (RFE) process

# S3 method for rfe
ggplot(data = NULL, mapping = NULL,
  metric = data$metric[1], output = "layered", ...,
  environment = NULL)

# S3 method for rfe
plot(x, metric = x$metric, ...)

Arguments

data	an object of class rfe.
mapping, environment	unused arguments to make consistent with ggplot2 generic method
metric	What measure of performance to plot. Examples of possible values are "RMSE", "Rsquared", "Accuracy" or "Kappa". Other values can be used depending on what metrics have been calculated.
output	either "data", "ggplot" or "layered". The first returns a data frame while the second returns a simple ggplot object with no layers. The third value returns a plot with a set of layers.
...	plot only: specifications to be passed to xyplot. The function automatically sets some arguments (e.g. axis labels) but passing in values here will over-ride the defaults.
x	an object of class rfe.

Value

a lattice or ggplot object

Details

These plots show the average performance versus the subset sizes.

Note

We using a recipe as an input, there may be some subset sizes that are not well-replicated over resamples. The `ggplot` method will only show subset sizes where at least half of the resamples have associated results.

References

Kuhn (2008), ``Building Predictive Models in R Using the caret'' (http://www.jstatsoft.org/article/view/v028i05/v28i05.pdf)

See also

rfe, xyplot, ggplot

Examples

if (FALSE) {
data(BloodBrain)

x <- scale(bbbDescr[,-nearZeroVar(bbbDescr)])
x <- x[, -findCorrelation(cor(x), .8)]
x <- as.data.frame(x)

set.seed(1)
lmProfile <- rfe(x, logBBB,
                 sizes = c(2:25, 30, 35, 40, 45, 50, 55, 60, 65),
                 rfeControl = rfeControl(functions = lmFuncs,
                                         number = 200))
plot(lmProfile)
plot(lmProfile, metric = "Rsquared")
ggplot(lmProfile)
}