TSDT (version 1.0.8)

parse_rpart: parse_rpart

Description

Extract splits from an rpart.object returned from a call to rpart().

Usage

parse_rpart(tree, include_subgroups = FALSE)

Value

A data.frame containing a parsed tree.

Arguments

tree

An rpart.object returned from call to rpart().

include_subgroups

A logical value indicating whether or not to include a string representation of the subgroups in the results. Defaults to FALSE.

Details

This function takes as its input an rpart.object returned from a call to rpart. It parses this rpart.object using rpart_nodes() and returns the splits in the tree. The data returned include the NodeID of the node to split, the NodeID of that node's parent, the NodeID of that nodes left child and right child, the number of observations in that node, the variable used in the split, the data type for the splitting variable, the logic indicating which observations will go to the node's left child, the value of the splitting variable at which the split ocurrs, the mean response value of the node, and (optionally) the string representation of the node's subgroup. A node's subgroup is defined by the sequence of splits from the root to that node.

See Also

rpart_nodes, rpart, rpart.object

Examples

requireNamespace( "rpart", quietly = TRUE )
## Generate example data containing response, treatment, and covariates
N <- 50
continuous_response = runif( min = 0, max = 20, n = N )
trt <- sample( c('Control','Experimental'), size = N, prob = c(0.4,0.6),
              replace = TRUE )
X1 <- runif( N, min = 0, max = 1 )
X2 <- runif( N, min = 0, max = 1 )
X3 <- sample( c(0,1), size = N, prob = c(0.2,0.8), replace = TRUE )
X4 <- sample( c('A','B','C'), size = N, prob = c(0.6,0.3,0.1), replace = TRUE )

## Fit an rpart model
fit <- rpart::rpart( continuous_response ~ trt + X1 + X2 + X3 + X4,
                     control = rpart::rpart.control( maxdepth = 3L ) )
fit

## Parse the results into a new data.frame
ex1 <- parse_rpart( fit, include_subgroups = TRUE )
ex1