splitstats: Stats about a tree's splits.

In itree: Tools for classification and regression trees, with an emphasis on interpretability.

Description

Computes some quantities about the variables used in the splits of an itree object. This is new to itree.

Usage

splitstats(tree,featlist=NULL)

Arguments

tree	an itree object.
featlist	A list of features to compute statistics of. Typically set to NULL, in which case we figure out the features using the itree object itself.

Value

A data frame with each row being a feature and each column being a statistic about how that feature is used in the tree.

Examples

require(mlbench); data(BostonHousing)
#fit a tree:
cart <- itree(medv~.,BostonHousing,minsplit=25,minbucket=25,cp=0)
splitstats(cart)

Example output

itree is based on the code of rpart.
Bug reports should be directed to this package's maintainer, not rparts'.

Loading required package: mlbench
       var split.ct total.inv.depth total.node.size isroot isdepth1
1     crim        2      0.11654527      0.10512821      0        0
2       zn        0      0.00000000      0.00000000      0        0
3    indus        0      0.00000000      0.00000000      0        0
4     chas        0      0.00000000      0.00000000      0        0
5      nox        1      0.03121748      0.02222222      0        0
6       rm        3      0.36420395      0.35769231      1        1
7      age        1      0.03642040      0.03547009      0        0
8      dis        3      0.14203954      0.12820513      0        0
9      rad        0      0.00000000      0.00000000      0        0
10     tax        0      0.00000000      0.00000000      0        0
11 ptratio        0      0.00000000      0.00000000      0        0
12       b        0      0.00000000      0.00000000      0        0
13   lstat        5      0.30957336      0.35128205      0        1

itree documentation built on May 2, 2019, 7:25 a.m.