processTree: Converts a tree to a graph

In burjandedes/randomForestDrawer: Interactive random forest model drawer

Description

Processes and converts a single tree of a random forest to an igraph graph.

Usage

processTree(treeGraphList, rawTreeDataFrame, defaultColor = "blue",
  defaultShape = "square", classes = NULL, colors = NULL, shapes = NULL)

Arguments

treeGraphList	A list which contains the graphs.
rawTreeDataFrame	A tree, which created with the randomForest::getTree function.
defaultColor	The default color of the nodes.
defaultShape	The deafult shape of the nodes.
classes	The classes of the random forest target variables (if any).
colors	The colors of the classes (if any).
shapes	The shapes of the classes (if any).

Details

With this function you can convert a tree of a random forest model to an igraph graph. At first we process the raw tree data and make a graph data from it. After that we add properties to the graph. And finally we append the new graph to a list which contains igraph graphs.

Value

A list which contains the new and a previous graphs.

Examples

## Not run: 
# Model must be a valid rf model with keep.forest=TRUE
treeGraphList <- list()
treeGraphList <- processTree(
                   treeGraphList
                   randomForest::getTree(model, k=1, labelVar=TRUE),
                 )
print(treeGraphList)

## End(Not run)

burjandedes/randomForestDrawer documentation built on May 18, 2019, 9:16 p.m.