\name{Control Forest Hyper Parameters}

\alias{cforest_control}

\alias{cforest_classical}

\alias{cforest_unbiased}

\title{ Control for Conditional Tree Forests }

\description{

  Various parameters that control aspects of the `cforest' fit via

  its `control' argument.

}

\usage{

cforest_unbiased(\dots)

cforest_classical(\dots)

cforest_control(teststat = "max",

                testtype = "Teststatistic",

                mincriterion = qnorm(0.9),

                savesplitstats = FALSE,

                ntree = 500, mtry = 5, replace = TRUE,

                fraction = 0.632, trace = FALSE, \dots)

}

\arguments{

  \item{teststat}{ a character specifying the type of the test statistic

                       to be applied. }

  \item{testtype}{ a character specifying how to compute the distribution of

                   the test statistic. }

  \item{mincriterion}{ the value of the test statistic (for \code{testtype == "Teststatistic"}),

                       or 1 - p-value (for other values of \code{testtype}) that

                       must be exceeded in order to implement a split. }

  \item{mtry}{ number of input variables randomly sampled as candidates 

               at each node for random forest like algorithms. Bagging, as special case 

               of a random forest without random input variable sampling, can 

               be performed by setting \code{mtry} either equal to \code{NULL} or 

               manually equal to the number of input variables.}

  \item{savesplitstats}{ a logical determining whether the process of standardized

                         two-sample statistics for split point estimate

                         is saved for each primary split.}

  \item{ntree}{ number of trees to grow in a forest.}

  \item{replace}{ a logical indicating whether sampling of observations is 

                 done with or without replacement.}

  \item{fraction}{ fraction of number of observations to draw without 

                   replacement (only relevant if \code{replace = FALSE}).}

  \item{trace}{ a logical indicating if a progress bar shall be printed

                while the forest grows.}

  \item{\dots}{ additional arguments to be passed to 

                \code{\link{ctree_control}}.}

}

\details{

  All three functions return an object of class \code{\link{ForestControl-class}}

  defining hyper parameters to be specified via the \code{control} argument

  of \code{\link{cforest}}.

  The arguments \code{teststat}, \code{testtype} and \code{mincriterion}

  determine how the global null hypothesis of independence between all input

  variables and the response is tested (see \code{\link{ctree}}). The 

  argument \code{nresample} is the number of Monte-Carlo replications to be

  used when \code{testtype = "MonteCarlo"}.

  A split is established when the sum of the weights in both daugther nodes

  is larger than \code{minsplit}, this avoids pathological splits at the

  borders. When \code{stump = TRUE}, a tree with at most two terminal nodes

  is computed.

  The \code{mtry} argument regulates a random selection of \code{mtry} input 

  variables in each node. Note that here \code{mtry} is fixed to the value 5 by 

  default for merely technical reasons, while in \code{\link[randomForest]{randomForest}} 

  the default values for classification and regression vary with the number of input 

  variables. Make sure that \code{mtry} is defined properly before using \code{cforest}.

  It might be informative to look at scatterplots of input variables against

  the standardized two-sample split statistics, those are available when

  \code{savesplitstats = TRUE}. Each node is then associated with a vector

  whose length is determined by the number of observations in the learning

  sample and thus much more memory is required.

  The number of trees \code{ntree} can be increased for large numbers of input variables.

  Function \code{cforest_unbiased} returns the settings suggested 

  for the construction of unbiased random forests (\code{teststat = "quad", testtype = "Univ", 

    replace = FALSE}) by Strobl et al. (2007)

  and is the default since version 0.9-90.

  Hyper parameter settings mimicing the behaviour of

  \code{\link[randomForest]{randomForest}} are available in

  \code{cforest_classical} which have been used as default up to

  version 0.9-14. 

  Please note that \code{\link{cforest}}, in contrast to 

  \code{\link[randomForest]{randomForest}}, doesn't grow trees of

  maximal depth. To grow large trees, set \code{mincriterion = 0}.

}

\value{

  An object of class \code{\link{ForestControl-class}}.

}

\references{

    Carolin Strobl, Anne-Laure Boulesteix, Achim Zeileis and Torsten Hothorn (2007).

    Bias in Random Forest Variable Importance Measures: Illustrations, Sources and  

    a Solution. \emph{BMC Bioinformatics}, \bold{8}, 25. 

    \url{http://www.BioMedCentral.com/1471-2105/8/25/}

}

\keyword{misc}