% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/var.relations.mfi.R
\name{var.relations.mfi}
\alias{var.relations.mfi}
\title{Investigate variable relations of a specific variable with mutual forest impact (corrected mean adjusted agreement).}
\usage{
var.relations.mfi(
  x = NULL,
  y = NULL,
  num.trees = 500,
  type = "regression",
  s = NULL,
  mtry = NULL,
  min.node.size = 1,
  num.threads = NULL,
  status = NULL,
  save.ranger = FALSE,
  create.forest = TRUE,
  forest = NULL,
  save.memory = FALSE,
  case.weights = NULL,
  variables,
  candidates,
  p.t = 0.01,
  select.rel = TRUE,
  method = "janitza"
)
}
\arguments{
\item{x}{data.frame of predictor variables with variables in
columns and samples in rows (Note: missing values are not allowed)}

\item{y}{vector with values of phenotype variable (Note: will be converted to factor if
classification mode is used). For survival forests this is the time variable.}

\item{num.trees}{number of trees. Default is 500.}

\item{type}{mode of prediction ("regression", "classification" or "survival"). Default is regression.}

\item{s}{predefined number of surrogate splits (it may happen that the actual number of surrogate splits differs in individual nodes). Default is 1 \% of no. of variables.}

\item{mtry}{number of variables to possibly split at in each node. Default is no. of variables^(3/4) ("^3/4") as recommended by (Ishwaran 2011). Also possible is "sqrt" and "0.5" to use the square root or half of the no. of variables.}

\item{min.node.size}{minimal node size. Default is 1.}

\item{num.threads}{number of threads used for determination of relations. Default is number of CPUs available.}

\item{status}{status variable, only applicable to survival data. Use 1 for event and 0 for censoring.}

\item{save.ranger}{set TRUE if ranger object should be saved. Default is that ranger object is not saved (FALSE).}

\item{create.forest}{set FALSE if you want to analyze an existing forest. Default is TRUE.}

\item{forest}{the random forest that should be analyzed if create.forest is set to FALSE. (x and y still have to be given to obtain variable names)}

\item{save.memory}{Use memory saving (but slower) splitting mode. No effect for survival and GWAS data. Warning: This option slows down the tree growing, use only if you encounter memory problems. (This parameter is transfered to ranger)}

\item{case.weights}{Weights for sampling of training observations. Observations with larger weights will be selected with higher probability in the bootstrap (or subsampled) samples for the trees.}

\item{variables}{variable names (string) for which related variables should be searched for (has to be contained in allvariables)}

\item{candidates}{vector of variable names (strings) that are candidates to be related to the variables (has to be contained in allvariables)}

\item{p.t}{p.value threshold for selection of related variables. Default is 0.01.}

\item{select.rel}{set False if only relations should be calculated and no related variables should be selected.}

\item{method}{Method  to  compute  p-values.   Use  "janitza"  for  the  method  by  Janitza  et  al. (2016) or "permutation" to utilize permuted relations.}
}
\value{
a list containing:
\itemize{
\item variables: the variables to which relations are investigated.
\item surr.res: a matrix with the mutual forest impact values with variables in rows and candidates in columns.
\item surr.perm: a matrix with the mutual forest impact values of the permuted variables with variables in rows and candidates in columns.
\item p.rel: a list with the obtained p-values for the relation analysis of each variable.
\item var.rel: a list with vectors of related variables for each variable.
\item ranger: ranger objects.
\item method: Method  to  compute  p-values: "janitza" or "permutation".
\item p.t: p.value threshold for selection of related variables


}
}
\description{
This function corrects the mean adjusted agreement by a permutation approach and generates the relation parameter mutual forest impact. Subsequently p-values are determined and related variables are selected.
}
\examples{
# read data
data("SMD_example_data")
x = SMD_example_data[,2:ncol(SMD_example_data)]
y = SMD_example_data[,1]
\donttest{
# calculate variable relations
set.seed(42)
res = var.relations.mfi(x = x, y = y, s = 10, num.trees = 100, variables = c("X1","X7"), candidates = colnames(x)[1:100])
res$var.rel[[1]]
}

}