FFTrees (version 1.1.3)

predict.fft: Applies an existing fft to a new dataset

Description

Applies an existing fft to a new dataset

Usage

"predict"(object = NULL, data = NULL, formula = NULL, which.tree = NULL, level.name.v = NULL, level.threshold.v = NULL, level.sigdirection.v = NULL, level.exit.v = NULL, level.class.v = NULL, ...)

Arguments

object
(M) An fft object created from the fft() function.
data
(M) An m x n dataframe containing n cue values for each of the m exemplars.
formula
a formula
which.tree
which.tree An integer indicating which tree to plot (only valid when the tree argument is non-empty). To plot the best training (or test) tree with respect to v (HR - FAR), use "best.train" or "best.test"
level.name.v
A character indicating the names of the levels in the tree separated by ;. For example "age;sex;occupation"
level.threshold.v
(M) A character indicating the level thresholds separated by ;. For example "25;female;occupation"
level.sigdirection.v
(M) A character vector of length n indicating the direction for which exemplars are classified as signals for each cue. Values must be in the set "<" (strictly="" less="" than),="" "<=" (less than or equal to), " ="="" (equal),="" "!=" (unequal), ">=" (greater than or equal to), or ">" (strictly greater than)/
level.exit.v
(B) A numeric vector of length n indicating the exit direction for each level. 0 = noise clasification, 1 = signal decision, .5 = both.
level.class.v
(B) A character vector of length n indicating the class of the cues for each level. "F" = factor, "N" = numeric, "L" = logical.
...
Additional arguments passed on to predict()

Value

A list of length 3. The first element "decision.df" is a dataframe with the decisions (and level of decisions) for each exemplar. The second element, "final.df" is a dataframe showing final tree accuracy statistics. The third element "level.df" shows tree accuracy statistics at each level.