Update or Re-fit a SA or GA Model

update allows a user to over-ride the search iteration selection process.

Based on the results of plotting a gafs or safs object, these functions can be used to supersede the number of iterations determined analytically from the resamples.

Any values of ... originally passed to gafs or safs are automatically passed on to the updated model (i.e. they do not need to be supplied again to update.

# S3 method for safs
update(object, iter, x, y, ...)

Arguments

object	An object produced by gafs or safs
iter	a single numeric integer
x, y	the original training data used in the call to gafs or safs. Only required for non-recipe methods.
...	not currently used

Value

an object of class gafs or safs

See also

gafs, safs

Examples

if (FALSE) {
set.seed(1)
train_data <- twoClassSim(100, noiseVars = 10)
test_data  <- twoClassSim(10,  noiseVars = 10)

## A short example
ctrl <- safsControl(functions = rfSA,
                    method = "cv",
                    number = 3)

rf_search <- safs(x = train_data[, -ncol(train_data)],
                  y = train_data$Class,
                  iters = 3,
                  safsControl = ctrl)

rf_search2 <- update(rf_search,
                         iter = 1,
                         x = train_data[, -ncol(train_data)],
                     y = train_data$Class)
rf_search2
}