xval {conTree}R Documentation

Cross-validate boosted contrast tree boosted with (new) data

Description

Cross-validate boosted contrast tree boosted with (new) data

Usage

xval(
  mdl,
  x,
  y,
  z,
  num = length(mdl$tree),
  del = 10,
  span = 0.15,
  ylab = "Average  Discrepancy",
  doplot = "first",
  doprint = FALSE,
  col = "red"
)

Arguments

mdl

model output from modtrast()

x

data predictor variables is same format as input to modtrast

y

data y values is same format as input to modtrast

z

data z values is same format as input to modtrast

num

number of trees used to compute model values

del

plot discrepancy value computed every del-th iteration (tree)

span

running median smoother span (doplot=TRUE, only)

ylab

graphical parameter ('doplot="first", only)

doplot

logical flag. doplot="first" implies start new display. doplot="next" implies super impose plot on existing display. doplot="none" implies no plot displayed.

doprint

logical flag TRUE/FALSE implies do/don't print progress while executing, default FALSE

col

color of plotted curve

Value

a named list of two items: ntree the iteration numbers, and error the corresponding discrepancy values

See Also

contrast()


[Package conTree version 0.3-1 Index]