RFSRCModel: Fast Random Forest (SRC) Model

In MachineShop: Machine Learning Models and Tools

Fast Random Forest (SRC) Model

Description

Fast OpenMP computing of Breiman's random forest for a variety of data settings including right-censored survival, regression, and classification.

Usage

RFSRCModel(
  ntree = 1000,
  mtry = integer(),
  nodesize = integer(),
  nodedepth = integer(),
  splitrule = character(),
  nsplit = 10,
  block.size = integer(),
  samptype = c("swor", "swr"),
  membership = FALSE,
  sampsize = if (samptype == "swor") function(x) 0.632 * x else function(x) x,
  nimpute = 1,
  ntime = integer(),
  proximity = c(FALSE, TRUE, "inbag", "oob", "all"),
  distance = c(FALSE, TRUE, "inbag", "oob", "all"),
  forest.wt = c(FALSE, TRUE, "inbag", "oob", "all"),
  xvar.wt = numeric(),
  split.wt = numeric(),
  var.used = c(FALSE, "all.trees", "by.tree"),
  split.depth = c(FALSE, "all.trees", "by.tree"),
  do.trace = FALSE,
  statistics = FALSE
)

RFSRCFastModel(
  ntree = 500,
  sampsize = function(x) min(0.632 * x, max(x^0.75, 150)),
  ntime = 50,
  terminal.qualts = FALSE,
  ...
)

Arguments

ntree	number of trees.
mtry	number of variables randomly selected as candidates for splitting a node.
nodesize	minumum size of terminal nodes.
nodedepth	maximum depth to which a tree should be grown.
splitrule	splitting rule (see rfsrc).
nsplit	non-negative integer value for number of random splits to consider for each candidate splitting variable.
block.size	interval number of trees at which to compute the cumulative error rate.
samptype	whether bootstrap sampling is with or without replacement.
membership	logical indicating whether to return terminal node membership.
sampsize	function specifying the bootstrap size.
nimpute	number of iterations of the missing data imputation algorithm.
ntime	integer number of time points to constrain ensemble calculations for survival outcomes.
proximity	whether and how to return proximity of cases as measured by the frequency of sharing the same terminal nodes.
distance	whether and how to return distance between cases as measured by the ratio of the sum of edges from each case to the root node.
forest.wt	whether and how to return the forest weight matrix.
xvar.wt	vector of non-negative weights representing the probability of selecting a variable for splitting.
split.wt	vector of non-negative weights used for multiplying the split statistic for a variable.
var.used	whether and how to return variables used for splitting.
split.depth	whether and how to return minimal depth for each variable.
do.trace	number of seconds between updates to the user on approximate time to completion.
statistics	logical indicating whether to return split statistics.
terminal.qualts	logical indicating whether to return terminal node membership information.
...	arguments passed to RFSRCModel.

Details

Response types:

factor, matrix, numeric, Surv

Automatic tuning of grid parameters:

mtry, nodesize

Default argument values and further model details can be found in the source See Also links below.

In calls to varimp for RFSRCModel, argument type may be specified as "permute" (default) for permutation of OOB cases, as "random" for permutation replaced with random assignment, or as "anit" for cases assigned to the split opposite of the random assignments. Variable importance is automatically scaled to range from 0 to 100. To obtain unscaled importance values, set scale = FALSE. See example below.

Value

MLModel class object.

See Also

rfsrc, rfsrc.fast, fit, resample

Examples

## Requires prior installation of suggested package randomForestSRC to run

model_fit <- fit(sale_amount ~ ., data = ICHomes, model = RFSRCModel)
varimp(model_fit, method = "model", type = "random", scale = TRUE)

MachineShop documentation built on Sept. 18, 2023, 5:06 p.m.