treeNormalise: Tree Normalise

In BlueTea88/xgboostcon: XGBoost conditions and parameter ranking

Description

Normalise tree conditions

Usage

treeNormalise(in.conditions, in.data, in.weights = NULL)

Arguments

in.conditions	feature conditions of trees, extracted using treeConditions()
in.data	training data used to fit the model in the form of a data.table
in.weights	vector of weights corresponding to each observation of in.data, defaults to a vector of 1s

Details

Normalise tree conditions by padding out lower order effects and setting the values of lower order conditions based on the weighted mean. The distribution of features is obtained from in.data and in.weights.

BlueTea88/xgboostcon documentation built on May 17, 2019, 11:05 a.m.